Biopsychology
a chill guy
BrainandBehaviorACognitiveNeurosciencePerspective.pdf.zipBrain and Behavior A Cognitive Neuroscience Perspective.pdf
# 158305 Cust: OUP Au: Eagleman Pg. No. i Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
A Cognitive Neuroscience Perspective
Brain and Behavior
00-Eagleman-FM.indd 1 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. ii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
00-Eagleman-FM.indd 2 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. iii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
DAVID EAGLEMAN Baylor College of Medicine, Department of Neuroscience Director, Initiative on Neuroscience and Law
JONATHAN DOWNAR Department of Psychiatry and Institute of Medical Science, University of Toronto Toronto Western Hospital, University Health Network
New York Oxfor d Oxfor d University Pr ess
A Cognitive Neuroscience Perspective
Brain and Behavior
00-Eagleman-FM.indd 3 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. iv Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide.
Oxford New York Auck land Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto
With offices in A rgentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungar y Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam
Copyright © 2016 by Oxford University Press
For titles covered by Section 112 of the US Higher Education Opportunity Act, please visit w w w.oup.com/us/he for the latest information about pricing and alternate formats.
Published by Oxford University Press 198 Madison Avenue, New York, N Y 10016 http://w w w.oup.com
Oxford is a registered trademark of Oxford University Press.
A ll rights reser ved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or other wise, without the prior permission of Oxford University Press.
Librar y of Congress Cataloging-in-Publication Data Eagleman, David. Brain and behavior : a cognitive neuroscience perspective / David Eagleman, Baylor College of Medicine, Department of Neuroscience, Director, Initiative on Neuroscience and Law, Jonathan Downar, Department of Psychiatr y and Institute of Medical Science, University of Toronto, Toronto Western Hospital, University Health Network. pages cm Includes bibliographical references and index. ISBN 978-0-19-537768-2 1. Cognitive neuroscience. 2. Neuropsychiatr y. I. Downar, Jonathan. II. Title. QP360.5.E24 2016 612.8'233—dc23
2015013925
Printing number: 9 8 7 6 5 4 3 2 1
Printed in the United States of A merica on acid-free paper
00-Eagleman-FM.indd 4 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. v Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
DEDICATION
From David: To Cirel, Arthur, Read, Francis, and Sarah, in the order
that I met you.
From Jonathan: In loving memory of Ann Downar—devoted mother of
two sons, and keen-eyed editor of many a manuscript—a gifted wielder of the red pen, from
whom no detail was too small to escape notice. With gratitude that you were able to see this book take shape
in its early days and with the hope that its final form would have made you proud.
00-Eagleman-FM.indd 5 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. vi Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
BRIEF CONTENTS PREFACE xv ABOUT THE AUTHORS xxi
PART I THE BASICS CHAPTER 1 Introduction 2 CHAPTER 2 The Brain and Nervous System 36 CHAPTER 3 Neurons and Synapses 74 CHAPTER 4 Neuroplasticity 102
PART II HOW THE BRAIN INTERACTS WITH THE WORLD CHAPTER 5 Vision 130 CHAPTER 6 Other Senses 162 CHAPTER 7 The Motor System 196
PART III HIGHER LEVELS OF INTERACTION CHAPTER 8 Attention and Consciousness 232 CHAPTER 9 Memory 270 CHAPTER 10 Sleep 308 CHAPTER 11 Language and Lateralization 336
PART IV MOTIVATED BEHAVIORS CHAPTER 12 Decision Making 362 CHAPTER 13 Emotions 398 CHAPTER 14 Motivation and Reward 438 CHAPTER 15 Social Cognition 472
PART V DISORDERS OF BRAIN AND BEHAVIOR CHAPTER 16 Neurological and Psychiatric Disorders 514
GLOSSARY 556 REFERENCES 585 CREDITS 637 NAME INDEX 643 SUBJECT INDEX 657
00-Eagleman-FM.indd 6 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. vii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The Payoffs of Cognitive Neuroscience 28 Healing the Disordered Brain 29 Enhancing Human Abilities 30 Blueprints for A rtificial Cognition 31 Brain-Compatible Social Policies 31
Conclusion 33 Key Principles 33 Key Terms 34 Review Questions 34 Critical-Thinking Questions 35
CHAPTER 2 The Brain and Nervous System 36
LE A R NING OBJECTI V ES 36 STA RTI NG OU T: The Brains of Creatures Great and
Small 38 A n Overview of the Nervous System 39
Why Put Your Neurons in a Brain at A ll? 39 The Common Features of Ever y Central Ner vous System 40 Getting Oriented in the Brain 42
The Peripheral Nervous System 43 Separate Systems for the Inner and Outer Environments 43 A Ner vous System with Segmental Organization 44
The Spinal Cord 47 Circuits within a Segment: Spinal Reflexes 47
Case Stu dy: Christopher Reeve, 1952–2004 47
Complex Circuits across Segments: Central Pattern Generators 49
The Bigger Pictu r e: In Search of a Cure for Spinal Cord Injury 50
The Brainstem 51 Medulla Oblongata and Pons 51
Neu roscience of Ev ery day Life: W hy Do We Get the Hiccups? 53
Midbrain 53 Most Cranial Ner ves Emerge from the Brainstem 54
The Cerebellum 54 Circuitr y of the “Little Brain” 56 Functions of the Little Brain 56
The Diencephalon: Hypothalamus and Thalamus 57 Hy pothalamus: A Keystone Structure in Homeostasis 57 Thalamus 59
Case Stu dy: Waking the Brain 61
PREFACE xv ABOUT THE AUTHORS xxi
PART I THE BASICS CHAPTER 1 Introduction 2
LE A R NING OBJECTI V ES 2 STA RTI NG OU T: A Spark of Awe in the Darkness 4
W ho A re We? 5 The Mission of Cognitive Neuroscience 5 Neuroscience Is a Relatively New Field 6
In Pursuit of Principles 6 The Functions behind the Form 6 Which Parts Matter? 7 What Is the Brain For? 8
How We K now W hat We K now 9 Connectional Methods 10 Correlational Methods 11
R esea rch M ethods: Magnetic Resonance Imaging 12
Lesion Methods 13 Stimulation Methods 14 A Toolbox of Complementar y Methods 16
Thinking Critically about the Brain 16 Is the Brain Equipped to Understand Itself ? 16 Biases and Pitfalls in Human Cognition 17 A Toolbox of Critical-Think ing Techniques 18
The Big Questions in Cognitive Neuroscience 19 Why Have a Brain at A ll? (Chapter 2) 19 How Is Information Coded in Neural Activity? (Chapter 3) 20 How Does the Brain Balance Stability against Change? (Chapter 4) 20 Why Does Vision Have So Little to Do with the Eyes? (Chapter 5) 21 How Does the Brain Stitch Together a Picture of the World from Different Senses? (Chapter 6) 21 How Does the Brain Control Our Actions? (Chapter 7) 22 What Is Consciousness? (Chapter 8) 22 How A re Memories Stored and Retrieved? (Chapter 9) 23 Why Do Brains Sleep and Dream? (Chapter 10) 24 How Does the Human Brain Acquire Its Unique Ability for Language? (Chapter 11) 24 How Do We Make Decisions? (Chapter 12) 25 What A re Emotions? (Chapter 13) 25 How Do We Set Our Priorities? (Chapter 14) 27 How Do I K now What You’re Think ing? (Chapter 15) 27 What Causes Disorders of the Mind and the Brain? (Chapter 16) 28
CONTENTS
00-Eagleman-FM.indd 7 29/10/15 2:55 pm
viii Contents
# 158305 Cust: OUP Au: Eagleman Pg. No. viii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Key Principles 100 Key Terms 100 Review Questions 101 Critical-Thinking Questions 101
CHAPTER 4 Neuroplasticity 102 LE A R NING OBJECTI V ES 102 STA RTI NG OU T: The Child with Half a Brain 104
The Brain Dynamically Reorganizes to Match Its Inputs 105 Changes to the Body Plan 105
Case Stu dy: Phantom Sensation 106 R esea rch M ethods: Mapping Out the Brain 107
Changes to Sensor y Input 107
The Brain Distributes Resources Based on Relevance 109 The Role of Behavior 109 The Role of Relevance: Gating Plasticity with Neuromodulation 109
Neu roscience of Ev ery day Life: Pianists and Violinists Have Different Brains 110
Case Stu dy: The Government Worker with the Missing Brain 112
The Brain Uses the Available Tissue 113 Maps Adjust Themselves to the Available Brain Tissue 113 Cortical Reorganization after Brain Damage 114
A Sensitive Period for Plastic Changes 114 A Window of Time to Make Changes 114
Case Stu dy: Danielle, the Feral Child in the Window 115
The Sensitive Period in Language 116 Neuromodulation in Young Brains 116
Hardwiring versus World Experience 117 Aspects of the Brain A re Preprogrammed 117 Experience Changes the Brain 117 Brains Rely on Experience to Unpack Their Programs Correctly 119
The Mechanisms of Reorganization 120 Neurons Compete for Limited Space 120 Competition for Neurotrophins 121 Rapid Changes: Unmask ing Ex isting Connections 122 Slow Changes: Grow th of New Connections 123
Changing the Input Channels 123 Case Stu dy: The Man W ho Climbs with His Tongue 125 The Bigger Pictu r e: Adding New Peripherals 126
Conclusion 126 Key Principles 127 Key Terms 127 Review Questions 128 Critical-Thinking Questions 128
The Telencephalon: Cerebral Cortex and Basal Ganglia 61 Cerebral Cortex 61 Basal Ganglia 64
R esea rch M ethods: Cytoarchitecture of the Cortex 64
Uniting the Inside and Outside Worlds 66 The Limbic System 66 The Ventricular System and Brain Function 68
Conclusion 69 Key Principles 71 Key Terms 71 Review Questions 72 Critical-Thinking Questions 73
CHAPTER 3 Neurons and Synapses 74 LE A R NING OBJECTI V ES 74 STA RTI NG OU T: The Kabuki Actor and the Pufferfish 76
The Cells of the Brain 77 Neurons: A Close-Up View 77 Many Different Ty pes of Neurons 79 Glial Cells 80
R esea rch M ethods: Visualizing Neurons and Their Products 81
Synaptic Transmission: Chemical Signaling in the Brain 83 Release of Neurotransmitter at the Synapse 83 Ty pes of Neurotransmitters 84 Receptors 85 Postsynaptic Potentials 86
The Bigger Pictu r e: Psychoactive Drugs 87
Spikes: Electrical Signaling in the Brain 88 Adding Up the Signals 88 How an Action Potential Travels 89 Myelinating A xons to Make the Action Potential Travel Faster 90 Action Potentials Reach the Terminals and Cause Neurotransmitter Release 91
Case Stu dy: Multiple Sclerosis 91 Neu roscience of Ev ery day Life: The Magic of a Local
Anesthetic 92
W hat Do Spikes Mean? The Neural Code 93 Encoding Stimuli in Spikes 93 Decoding Spikes 95
R esea rch M ethods: Recording Action Potentials with Electrodes 96
Individuals and Populations 97 Populations of Neurons 97 Forming a Coalition: What Constitutes a Group? 98 Open Questions for Future Investigation 99
Conclusion 100
00-Eagleman-FM.indd 8 29/10/15 2:55 pm
Contents ix
# 158305 Cust: OUP Au: Eagleman Pg. No. ix Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
CHAPTER 6 Other Senses 162 LE A R NING OBJECTI V ES 162 STA RTI NG OU T: The Man with the Bionic Ear 164
Detecting Data from the World 165 Hearing 165
R esea rch M ethods: Psychophysics 166
The Outer and Middle Ear 167 Converting Mechanical Information into Electrical Signals: The Inner Ear 168
Neu roscience of Ev ery day Life: The Undetectable Cell Phone 169
The Auditor y Ner ve and Primar y Auditor y Cortex 170 The Hierarchy of Sound Processing 171 Sound Localization 172 Balance 173
The Somatosensory System 174 Touch 174 Temperature 175 Pain 175
Case Stu dy: The Pain of a Painless Existence 176
Proprioception 177 Interoception 177 The Somatosensor y Pathway 179
Chemical Senses 180 Taste 180 Smell 183 The Sense of Flavor 184 Pheromones 185
The Brain Is Multisensory 185 Synesthesia 186 Combining Sensor y Information 186 The Binding Problem 188 The Internal Model of the World 189
Case Stu dy: The Paralyzed Supreme Court Justice W ho Claimed He Could Play Football 189
Time Perception 190 Conclusion 193 Key Principles 193 Key Terms 194 Review Questions 194 Critical-Thinking Questions 195
CHAPTER 7 The Motor System 196 LE A R NING OBJECTI V ES 196 STA RTI NG OU T: “‘Locked-In Syndrome”’ 198
Muscles 199 Skeletal Muscle: Structure and Function 199 The Neuromuscular Junction 200
PART II HOW THE BRAIN INTERACTS WITH THE WORLD
CHAPTER 5 Vision 130 LE A R NING OBJECTI V ES 130 STA RTI NG OU T: Vision Is More Than the Eyes 132
Visual Perception 132 What Is It Like to See? 132 Signal Transduction 133
A natomy of the Visual System 134 Sensor y Transduction: The Eye and Its Retina 134
Case Stu dy: The Bionic Retina 137
Path to the Visual Cortex: The Lateral Geniculate Nucleus 139 The Visual Cortex 139 Two Eyes A re Better Than One: Stereo Vision 141
Neu roscience of Ev ery day Life: Random-Dot Stereograms 142
Higher Visual A reas 142 Secondar y and Tertiar y Visual Cortex: Processing Becomes More Complex 142 Ventral Stream: What an Object Is 143
The Bigger Pictu r e: Reading the Movies in Our Minds 145
Dorsal Stream: How to Interact with the World 146 Case Stu dy: The World in Snapshots 147
Attention and the Dorsal Stream 148 Comparing the Ventral and Dorsal Processing Streams 149 The Bigger Picture of the Visual Brain 150
Case Stu dy: The Blind Woman W ho Could See, Sort Of 150
Perception Is Active, Not Passive 151 Interrogating the Scene with Our Eyes 151 The Blind Spot 152 Seeing the Same Object Different Ways: Multistability 152 Binocular R ivalr y: Different Images in the Two Eyes 152 We Don’t See Most of What Hits Our Eyes: Fetching Information on a Need-to-K now Basis 153
Vision Relies on Expectations 154 Change Blindness 154 Saving Resources by Embedding Prior Experience 155 Unconscious Inference 156 Activity from Within 157 Feedback A llows an Internal Model 157
Conclusion 158 Key Principles 159 Key Terms 160 Review Questions 160 Critical-Thinking Questions 161
00-Eagleman-FM.indd 9 29/10/15 2:55 pm
x Contents
# 158305 Cust: OUP Au: Eagleman Pg. No. x Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Inattentional Blindness 236 Neu roscience of Ev ery day Life: Stage Magic 237
Approaches to Studying Attention and Awareness 238 Attentional Orienting Paradigms: A iming the “Spotlight” of Attention 238 The Oddball Paradigm: Monitoring a Physiological Measure of Attention 239 Uncoupling Sensor y Input from Perception: Sensor y R ivalr y 240
Neural Mechanisms of Attention and Awareness 241 Seek ing the Correlates of Consciousness 241 Hemineglect: A Disorder of Attention and Awareness 242
Case Stu dy: Unaware of Half of the World 243
Neural Correlates of Attention: A Single Network or Many? 245
Case Stu dy: W hose Arm Is This, Anyway? 247
Sites of Attentional Modulation: Neurons and Neural Populations 247 The Biased-Competition Model of Attention 247 Attention and Single Neurons: Enhancing the Signal 248 Attention and Local Groups of Neurons 250
Synchronization, Attention, and Awareness 250 Coma and Vegetative State: A natomy of the Conscious State 253 Why Should Synchronization Matter? 253 Unconsciousness: Coma and Vegetative State 254
Case Stu dy: Waking the Brain 255
Midbrain and Thalamus: Key Players in the Conscious State 256
A nesthesia and Sleep: R hythms of Consciousness 258 Sleep: Unraveling the R hythm of Consciousness 258 A nesthesia: Reversible, A rtificial Unconsciousness 260
Theories of Consciousness 262 Dualism: The Mind–Body Problem 262 Functionalist Theories of Consciousness 263 Consciousness and the Integration of Information 265
Conclusion 266 Key Principles 267 Key Terms 267 Review Questions 268 Critical-Thinking Questions 269
CHAPTER 9 Memory 270 LE A R NING OBJECTI V ES 270 STA RTI NG OU T: The Woman W ho Cannot Forget 272
The Many K inds of Memory 273 Work ing and Long-Term Memor y 273
The Spinal Cord 201 Lower Motor Neurons 201 Spinal Motor Circuits: Reflexes 202 Spinal Motor Circuits: Central Pattern Generators 203 Descending Pathways of Motor Control 204
The Cerebellum 205 The Circuitr y of the Cerebellum 206 Motor Functions of the Cerebellum 208 Nonmotor Functions of the Cerebellum 208
The Motor Cortex 209 Motor Cortex: Neural Coding of Movements 211 Motor Cortex: Recent Controversies 212
The Bigger Pictu r e: Neural Implants for Motor Control 214
The Prefrontal Cortex: Goals to Strategies to Tactics to Actions 215 The Functional Organization of the Prefrontal Cortex in Motor Control 215 Sensor y Feedback 216 Mirror Neurons in Premotor Cortex 217 Control Stages of the Motor Hierarchy 218
Basal Ganglia 219 Components of the Basal Ganglia 219 Circuitr y of the Basal Ganglia 220 Diseases of the Basal Ganglia 220
Medial and Lateral Motor Systems: Internally and Externally Guided Movement Control 222 Organization of Medial Motor A reas 222 Functions of Medial and Lateral Motor Systems 222
Neu roscience of Ev ery day Life: W hy Can’ t I Multitask? 225
Did I Really Do That? The Neuroscience of Free Will 226
R esea rch M ethods: Neurosurgical Stimulation 228 Case Stu dy: Alien Hand Syndrome 229
Conclusion 229 Key Principles 230 Key Terms 230 Review Questions 231 Critical-Thinking Questions 231
PART III HIGHER LEVELS OF INTERACTION
CHAPTER 8 Attention and Consciousness 232
LE A R NING OBJECTI V ES 232 STA RTI NG OU T: The Stream of Consciousness 234
Awareness Requires Attention 235 Change Blindness 235
00-Eagleman-FM.indd 10 29/10/15 2:55 pm
Contents xi
# 158305 Cust: OUP Au: Eagleman Pg. No. xi Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The Limitations of Neural Networks 302 Neural Networks: Solving the Wrong Problem? 303 Remembering Relationships, Not Features 303 The Future of Memor y Research 303
Conclusion 304 Key Principles 305 Key Terms 305 Review Questions 306 Critical-Thinking Questions 307
CHAPTER 10 Sleep 308 LE A R NING OBJECTI V ES 308 STA RTI NG OU T: Caught between Sleeping and
Waking 310 Sleep and the Brain 311
The Brain Is Active during Sleep 311 R esea rch M ethods: Electroencephalography 312
The Neural Networks of Sleep 314 The Brain during R EM Sleep 314
The Circadian R hythm 316 Entrainment of the Circadian R hythm by Light Cues 316 The Circadian R hythm Is Not Fixed 317
Case Stu dy: The Shifted Circadian Rhythm 318
The Circadian R hythm and Napping 319 The Bigger Pictu r e: Schools and Circadian Rhythms 320
W hy Do Brains Sleep? 320 Four Theories of Sleeping: Restoration, Sur vival, Simulation, Learning 320 Rehearsal 321 Forgetting 323 Insight and the Restructuring of Information 324
Dreaming 324 Dream Content 325
Neu roscience of Ev ery day Life: Lucid Dreaming 326
Can Dreams Shed Light on Consciousness? 327 Dreams of the Future and How to Study Them 327
Sleep Deprivation and Sleep Disorders 328 Sleep Deprivation 328
Case Stu dy: Staying Awake 329
Insomnia 330 Hy persomnia 331
Case Stu dy: The Family W ho Couldn’ t Sleep 332
Parasomnias 333 Conclusion 333 Key Principles 334 Key Terms 335 Review Questions 335 Critical-Thinking Questions 335
Implicit Memor y 274 Explicit Memor y 275
Travels in Space and Time: The Hippocampus and Temporal Lobe 277
Case Stu dy: Gone but Not Forgotten: Henry Molaison, 1926–2008 277
A Map of the Medial Temporal Lobe 278 Episodic Memor y 279 Spatial Memor y 280 Theories of Hippocampal Function 281 Unif ying the Functions of the Hippocampus 282
Remembering the Future: Prospection and Imagination 282 How We Imagine Future Experiences 282
R esea rch M ethods: Localizing Human Brain Function 283
The Circuitr y of Prospection and Recollection 284 Neu roscience of Ev ery day Life: Simonides and the
Champions of Memory 285
Prospection in Other Species 286 Models of Prospection 287
The Confabulation of Reality 288 Confabulation in the Injured Brain 288
Case Stu dy: The Woman with a Thirty-Year-Old Baby 288
The A natomy of Spontaneous Confabulation 289 Confabulation in the Normal Brain 290 The A natomy of a False Memor y 291
The Bigger Pictu r e: Scanning for the Truth 292
The Mechanisms of Memory 292 General Mechanisms of Learning and Memor y 292 Memor y as Synaptic Change 293 Long-Term Potentiation and Depression of Synaptic Connections 293 The NMDA Receptor 293 Consolidation and Reconsolidation 294 Associative Neural Networks 295
Beyond Synaptic Plasticity: The Frontiers of Memory Mechanisms 296 Whole Neurons as a Substrate for Memor y? 296 New Neurons for New Memories 297 Spines: A nother Structural Basis for Memor y? 298 Look ing inside the Cell: Memor y in Chemical Reactions 298
Case Stu dy: The Flies with Photographic Memory 299
Epigenetics: Mak ing a Single Genome Play Different Tunes 300
The Mysteries of Memory 300 A re the Roles of LTP and LTD Overstated? 301 The Timing of Spikes 301
00-Eagleman-FM.indd 11 29/10/15 2:55 pm
xii Contents
# 158305 Cust: OUP Au: Eagleman Pg. No. xii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
W here Do Our Irrational Decisions Come From? 371 Decision Mak ing in Other Species 372 Do Irrational Decisions Come from Irrational People? 373 One Brain, Two Systems 373
How the Brain Decides 374 The Neural Mechanisms of Delay Discounting 374 Neural Mechanisms of Decisions under R isk 375 The Neural Basis of the Endowment Effect 377 The Neural Basis of the Framing Effect 378
The Common Currency of Subjective Value 379 Comparing Apples to Oranges 379
R esea rch M ethods: Charting the Landscape of Subjective Value 380
A Consistent Neural Basis for Subjective Value 380 Evaluation and the Orbitofrontal Cortex 381 One Currency, but Many Markets 382
Neu roscience of Ev ery day Life: Snack Food or Brussels Sprouts? 383
A Hierarchy of Internally Guided Decision Making 384 Internally and Externally Guided Decision Mak ing 384 Values into Goals 385 Goals into Plans 387 Plans into Behavior and Action 388
Modulators of Decision Making 389 Strategic Use of Decision-Mak ing Systems 389 Neurotransmitter Effects on Decision Mak ing 391
The Bigger Pictu r e: How to Avoid the Scorpion’s Sting 394
Conclusion 395 Key Principles 396 Key Terms 396 Review Questions 397 Critical-Thinking Questions 397
CHAPTER 13 Emotions 398 LE A R NING OBJECTI V ES 398 STA RTI NG OU T: Sadness, at the Flip of a Switch 400
Early Theories of Emotion 401 Emotional Expressions: Signposts on a Landscape of Inner States 401 The James–Lange Theor y of Emotion: A Bottom-Up Theor y 402 The Cannon–Bard Theor y: A Top-Down Theor y 404
Case Stu dy: Pathological Laughter and Crying 405
Two-Factor Theories: Reconciling Central and Peripheral Influences on Emotion 406
CHAPTER 11 Language and Lateralization 336
LE A R NING OBJECTI V ES 336 STA RTI NG OU T: The Stuttering King 338
Speech, Language, and Communication 338 Aphasia: The Loss of Language 339
Case Stu dy: The Woman W ho Couldn’ t Find Her Words 340
Broca’s Aphasia 341 Wernicke’s Aphasia 341
Case Stu dy: The Woman W ho Makes Up Words 342
A Language Network 343 The Larger Picture of Language-Specific Regions 344 Dyslex ia 347 Stuttering 348
Lateralization: The Two Hemispheres A re Not Identical 348 Tests for Dominance 349 Aprax ia 349 Hemispheric Differences 349 Two Brains in One? The Case of the Split-Brain Patients 350 Think ing about Cerebral Asymmetr y 352
Development of Language 352 Learning Language from Experience 353 Innate Language Tendencies 354 Socially and Emotionally Directed Learning 356
R esea rch M ethods: The Baby with No Privacy 357
Conclusion 359 Key Principles 359 Key Terms 360 Review Questions 360 Critical-Thinking Questions 361
PART IV MOTIVATED BEHAVIORS CHAPTER 12 Decision Making 362
LE A R NING OBJECTI V ES 362 STA RTI NG OU T: A Fatal Mistake, at the Highest
Place on Earth 364 How Do We Decide W hat to Do? 365
The Scorpion and the Frog 365 The Search for a “Physics” of Human Decisions 366 Homo economicus and Rational Choice Theor y 366
The Predictably Irrational Homo sapiens 367 Homo sapiens versus Homo economicus 367 Confused by Uncertainty 368 The Framing Effect and the Endowment Effect 369 The Illusor y Value of Procrastination 370
00-Eagleman-FM.indd 12 29/10/15 2:55 pm
Contents xiii
# 158305 Cust: OUP Au: Eagleman Pg. No. xiii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Midbrain Dopamine Neurons and the Common Currency of Motivation 447
Reward, Learning, and the Brain 447 Defining Reward 448 Learning from Reward Using Prediction Error 448 “Lik ing” Is Different from “Wanting” 450
Opioids and the Sensation of Pleasure 451 Opioids, Opioid Receptors, and Opioid Functions 451 Opioids and Reward 453
Dopamine, Learning, Motivation, and Reward 453 Dopamine Functions in Motivation and Reward 454 Unif ying the Functions of Dopamine 455
R esea rch M ethods: Measuring Neurotransmitter Levels in the Brain 456
Neurotransmitters A re Messengers, Not Functions 457
Addiction: Pathological Learning and Motivation 457 Addictive Substances Have Distorted Reward Value 457
Neu roscience of Ev ery day Life: The Pursuit of Happiness 458
Addiction Is a Result of Pathological Learning 460 The Circuitr y and Chemistr y of Addiction 461
Unlearning Addiction 462 The Challenge of Treatment 462
Case Stu dy: Pathological Gambling in a Patient with Parkinson’s Disease 463
Ex isting Approaches to Treatment 464 Future Approaches to Treatment 465
The Bigger Pictu r e: Finding the Motivation to Change 468
Conclusion 469 Key Principles 469 Key Terms 470 Review Questions 471 Critical-Thinking Questions 471
CHAPTER 15 Social Cognition 472 LE A R NING OBJECTI V ES 472 STA RTI NG OU T: W hy Risk Your Life for a Yellow
T-shirt? 474 Social Perception 475
What’s in a Face? 475 Do I Look Like a Liar to You? 476
Neu roscience of Ev ery day Life: A Poker Face 476
Social K nowledge and the Temporal Pole 478 Social Signals and the Superior Temporal Sulcus 479
Social Thinking: Theory of Mind 480 What Is Theor y of Mind? 480 Neural Mechanisms of Theor y of Mind 481
Core Limbic Structures: A mygdala and Hypothalamus 408 Hy pothalamus: Internal States, Homeostatic Drives 409
Case Stu dy: An Internal Growth of Rage 411
Do Hy pothalamic Circuits Generate Inner Emotional Experiences? 412 A mygdala: Externally Generated States and Drives 412 The A mygdala and Emotional Experience 413
Case Stu dy: The Woman W ho Knows No Fear 414
Hippocampus: Emotional Memories 416 Ventral Striatum: Pleasure and Reward 417 Bringing It A ll Together: The Circuit of Papez and the R ing of Limbic Cortex 418
The Bigger Pictu r e: The Ethics of Brain Stimulation in Human Beings 419
The Limbic Cortex and Emotions 420 The Interoceptive Insula: The “Feeling” Side of Emotions 420 Cingulate Cortex: A Motor Cortex for the Limbic System 421
Neu roscience of Ev ery day Life: Mental Effort 422
Ventromedial Prefrontal Cortex: A Generator of Gut Feelings 423
Limbic Association Cortex: Modulation of Emotion 425 The Mechanisms of Emotional Reappraisal 425 Brain Injur y, Brain Stimulation, and Emotion Regulation 426
Neurochemical Influences on Emotion 428 Case Stu dy: A Cure Born of Desperation 429
Serotonin and Mood 430 Norepinephrine and Mood 431 GA BA and A nx iety 432
Conclusion 433 Key Principles 435 Key Terms 435 Review Questions 436 Critical-Thinking Questions 437
CHAPTER 14 Motivation and Reward 438 LE A R NING OBJECTI V ES 438 STA RTI NG OU T: “More Important Than Survival
Itself ” 440 Motivation and Survival 440
Addiction: A n Illness of Motivation 440 Why Motivation Matters 441 Feelings: The Sensor y Side of Motivation 442
The Circuitry of Motivation: Basic Drives 443 Hy pothalamus and Homeostatic Drives 443 A mygdala and External-World Drives 445
00-Eagleman-FM.indd 13 29/10/15 2:55 pm
xiv Contents
# 158305 Cust: OUP Au: Eagleman Pg. No. xiv Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Huntington’s Disease: A Genetic Rarity, in Two Senses 524
Tourette Syndrome: A Case of Involuntary Volition? 527
Obsessive–Compulsive Disorder: Neurological or Psychiatric? 529
R esea rch M ethods: Voxel-Based Morphometry 532
Schizophrenia: A Dementia of the Young 534 Bipolar Disorder 540 Depression: A Global Burden 544
Impact of Depression 544 Case Stu dy: A Lifetime Studying , and Living with, Bipolar
Disorder 545
Causes of Depression 546 Neurochemical Effects of Depression on the Brain 548 Functional Effects of Depression on the Brain 549 Treatment of Depression 550
Conclusion 552 Key Principles 553 Key Terms 554 Review Questions 554 Critical-Thinking Questions 555
GLOSSARY 556 REFERENCES 585 CREDITS 637 NAME INDEX 643 SUBJECT INDEX 657
Mirror Neurons and Theor y of Mind 483 Disorders of Theor y of Mind 484
Social Feelings: Empathy and Its Many Components 487 A n Emotional Theor y of Mind 487 Empathy, Sympathy, and Compassion 488 Neural Mechanisms of Emotional Mimicr y and Contagion 489 Neural Mechanisms of Empathy, Sympathy, and A ntipathy 490 Disorders of Empathy 491
Social Emotions, Motivations, and Behavior 493 Social Emotions from Theor y of Mind 493
Case Stu dy: Acquired Sociopathy 493
Social Emotions from Social Values 495 Social Reward and Social Aversion 496 The A natomy of a Lie 498
Neurotransmitters and Social Behavior 499 R esea rch M ethods: Transcranial Direct Current
Stimulation 500
A n A ncient and Fundamental System 501 Ox ytocin 501 Vasopressin 503
The Bigger Pictu r e: The Brave New World of the “Cuddle Hormone”? 504
The Social Self 505 The Wondrous Self-Awareness of the Human Brain 505 Forms of Self-Awareness 505 Why Bother with Self-Awareness? 506 Neural Correlates of Self-Awareness 507 Disorders of Self-Awareness 508 Self-Awareness and Social Cognition 510
Case Stu dy: The Man in the Mirror 510
Conclusion 511 Key Principles 511 Key Terms 512 Review Questions 512 Critical-Thinking Questions 513
PART V DISORDERS OF BRAIN AND BEHAVIOR
CHAPTER 16 Neurological and Psychiatric Disorders 514
LE A R NING OBJECTI V ES 514 STA RTI NG OU T: Epilepsy: “The Sacred Disease” 516
A lzheimer’s Disease: Burning Out with Age? 517 Frontotemporal Dementia: Like a Cancer of the Soul 521
Case Stu dy: Ravel and “Bolero” 523
00-Eagleman-FM.indd 14 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xv Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Approach Brain and Behavior is a new kind of textbook for the emerg- ing field of cognitive and behavioral neuroscience. No other textbook tells the story of the brain in such a logical and meaningful way. Through the use of overarching principles rather than lists or facts, Brain and Behavior highlights what we understand about the function of the brain, as well as what we have left to learn and future directions.
Brain and Behavior illustrates current thinking in the field and builds scaffolding for you to learn new concepts. Without compromising important ideas, it covers a wide swath of territory critical for understanding the brain, from the basics of the nervous system to sensory and motor sys- tems, the frontal lobes, sleep, language, memory, drug addic- tion, and brain disorders. Throughout the book, the narrative emphasizes the dynamically changing nature of the brain (neuroplasticity) using clear and vibrant writing and fasci- nating real-life examples and applications.
Brain and Behavior presents the concepts of cognitive neuroscience as thoroughly as possible, using an easy and ac- cessible style that does not presuppose advanced knowledge. It features the following: • A principles-based approach. Students of all ages, and
especially undergraduates, find themselves frustrated with lists of unrelated facts to memorize. Overarching principles enable you to wrap your head around the big picture and learn how to mine for further details.
• A progressive structure. This book unfolds logically, be- ginning with the basics of the nervous system before moving to the brain’s interaction with the world (sensory and motor systems) and then to more complex interac- tions (attention, learning, sleeping, and dreaming). Building on this foundation, the book introduces still more complex interactions (language, decisions, emo- tions, motivation, and reward), before exploring the ways that the system can go awry (drug addiction, mental dis- orders, and neurological disorders).
• Engaging features. Several tools and features throughout each chapter help in the preparation for exams and high- light real-life examples and applications of the material.
Chapter Opening Each chapter opens with a rendering of the human body, emphasizing a key aspect related to the chapter topic, and includes a list of the major sections, features, and learning objectives to be covered.
The human brain is the most complex object we have found in the universe. There are more connections in a cubic milli- meter of neural tissue than there are stars in the Milky Way galaxy. So it is no surprise that even in the glow of remarkable advancement in recent decades, we find ourselves squinting to find the lay of the land. Even for experts in the field, the brain’s complexity can feel daunting at the best of times.
With this point in mind, we set out to write a cognitive neuroscience textbook that would help readers make sense of this complexity by focusing on fundamental scientific principles, patterns, and ways of thinking. Throughout the text, we prize understanding integration of principles over simple memorization of brain structures and scattered find- ings. As students from all backgrounds become increasingly interested in the brain, we wanted to capture the state of the science while distilling the expansive territory into under- standable parts.
Brain and Behavior covers a wide swath of territory criti- cal for understanding the brain, from the basics of the ner- vous system to sensory and motor systems, sleep, language, memory, emotions and motivation, social cognition, and brain disorders. Throughout the narrative we have sought to emphasize the dynamically changing nature of the brain through the mechanisms of neuroplasticity. In addition, wherever possible, we make reference to elements of neuro- science that are encountered in everyday life. We illustrate key points and concepts using case studies of rare but illumi- nating brain disorders. Brain and Behavior pulls together the best of our current knowledge about the brain while ac- knowledging our current areas of ignorance and pointing the reader toward our most promising directions for future research.
Brain and Behavior aims to present key concepts as thor- oughly as possible, in a reader-friendly style that does not presuppose advanced knowledge of the field. Our intention was to make the topic as accessible as possible to a wide un- dergraduate audience. However, it is our hope that students at all levels, and in other fields, will find this text to be a help- ful introductory guide to the complexities of the human brain.
W hether you are reading this book as an aspiring neuro- scientist or whether you are reading it simply as a fellow human being who wishes to better understand the miniature universe we carry inside our heads, we hope that you will come away from Brain and Behavior having gained a better understanding both of the human brain and of the human experience.
With best wishes, David Eagleman and Jonathan Downar
PREFACE
00-Eagleman-FM.indd 15 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xvi Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
xvi Preface
Inside the Chapter CASE STU DIES interwoven with the text present fascinat- ing human-interest stories that illustrate key content. These clinical cases include a woman incapable of feeling fear and a blind mountain climber who “sees” via electrical signals on his tongue.
Changing the Input Channels 125
# 158305 Cust: OUP Au: Eagleman Pg. No. 125 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
we think of as a visual cortex; plug an auditory stream into it, and it will become an auditory cortex.
As we have seen so far, the brain has a remarkable capac- ity to reconfigure itself in the face of new inputs, outputs, neural gains, or neural losses. That versatility opens the door to technologies that can deliver information to the brain through unusual sensory channels. For example, what if a blind person used the data stream from a video camera and converted it into sounds in her headphones? Would she eventually be able to see the world by listening? Welcome to the world of sensory substitution experiments, in which a deficient sensory channel is circumvented in favor of other routes to the brain (Lenay, Hanneton, Marque, & Genouel, 2003; Maidenbaum, Abboud, & Amedi, 2013; Poirier, De Volder, & Scheiber, 2007).
W hy does the BrainPort use the tongue, of all places? A lthough we normally think of the tongue as a taste organ, it also has the finest sense of touch on the entire body. It can distinguish stimuli only 1.6 mm apart—roughly twice as fine as a ty pical fingertip (Wilson, Walton, Tyler, & Wil- liams, 2012). This makes the tongue a great site for passing on new k inds of information, even learning to “see.” A grid of electrodes, the size of a postage stamp, zaps the tongue, converting the lattice of video pixels into “pixels” in the mouth (Bach-y-R ita, 2004; Bach-y-R ita, Collins, Saunders,
W hite, & Scadden, 1969). With practice, the tongue learns to interpret the signals that correspond to the visual prop- erties, such as how large an object is, how far away it is, and whether it’s moving in a particular direction. With the BrainPort, blind users can learn to navigate complex ob- stacle courses and throw balls into buckets. For sighted people, the BrainPort can be used to see in the dark . How is any of this possible? Vision isn’t about the eyes. It’s about the brain (K hoo, Seidel, & Zhu, 2012). Other sen- sor y substitution dev ices for the blind convert v ideo streams into patterns of touch on the lower back, sound for the ears, or small electric shocks to the sk in of the fore- head. Similarly, a sensor y substitution dev ice for the deaf uses a vest covered in v ibrator y motors to translate sound into patterns on the sk in (Nov ich & Eagleman, under rev iew). These amazing substitutions are possible only be- cause the brain can dy namically shape itself around what- ever input is presented. It even seems possible that in the near f uture people w ill feed information streams directly into their cortex.
In conclusion, the rerouting of information and the suc- cess of sensory substitution underscores the dynamic plas- ticity of brains. The principles of competition constantly reorganize the circuitry to optimize their representation of the input.
CASE STUDY: The Man Who Climbs with His Tongue Eric Weihenmayer is a mountaineer who has scaled Mount Everest—a feat made even more impressive by the fact that he is blind. As a child, Eric progressively lost his vision to a rare eye disease called retinoschi- sis, and he was rendered entirely blind by the age of 13. But that didn’t slow his ambition to become a climber. Given his condition, it’s cap- tivating to watch Eric scale shear rock faces, holding on to small crev- ices and protrusions. How does he know where to reach next? How does he do it?
Eric climbs with an electrode grid in his mouth called the BrainPort (FIGURE 4.26). The grid delivers little impulses to his tongue that mirror
the visual signals from a camera at- tached to his forehead. Eric reports that he first had to think hard about how the tongue stimulation might translate into edges and shapes. But
he learned, eventually, to recognize the stimulation as direct perception (Levy, 2008). He is now able to use the device for a low-resolution but effec- tive sense of his visual surroundings.
FIGURE 4.26 BrainPort. The BrainPort converts a video feed to corresponding electrical activity on the tongue. With this technology, blind users can come to understand their visual surroundings with high accuracy.
04-Eagleman_Chap04.indd 125 09/10/15 2:29 AM
Three special features appear throughout each chapter to highlight key themes and concepts:
NEUROSCIENCE OF EV ERY DAY LIFE explains how neuroscience directly relates to our daily lives, such as why people have difficulty multitasking.
Medial and Lateral Motor Systems: Internally and Externally Guided Movement Control 225
# 158305 Cust: OUP Au: Eagleman Pg. No. 225 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
NEUROSCIENCE OF EVERYDAY LIFE: Why Can’t I Multitask? Why is it so hard to do more than one thing at once? Why should doing one task interfere with doing another? Our mouths and lips are not neces- sary to drive a car, so why is it dan- gerous to talk on a hands-free cell phone while driving? When we look at a face, we don’t need to examine each part of the face one at a time to tell who we are looking at. We simply do it all at once. Why can’t we per- form many behaviors at once, in the same way that we perform many sensory tasks at once?
Neuroimaging studies have begun to search for the apparent “bottleneck” in information pro- cessing that limits our ability to mul- titask (FIGURE 7.28). One study (Dux, Ivanoff, Asplund, & Marois, 2006) had subjects perform two tasks almost simultaneously: pressing buttons according to an auditory cue and pronouncing syllables accord- ing to a visual cue. When these tasks occurred close together in time, the subjects’ reaction times began to slow: the limits of their multitasking abilities had been reached.
Functional MRI revealed a possi- ble basis for the slowing. Some brain regions were active for only one of these two tasks. However, other areas responded for both tasks; these areas could be poten- tial sites of the multitasking bottle- neck. Two of these regions showed responses that reflected the degree of reaction-time slowing (Dux et al., 2006). The posterior lateral prefron- tal cortex and, to some extent, the SMA both showed this apparent slowing of response activity. For competition between these two simple cue–response tasks, the bottleneck appeared to be in areas low in the motor hierarchy. However,
multitasking at higher levels of cog- nitive control may have a different basis. For example, compared with healthy individuals, patients with damage to the frontopolar cortex have trouble multitasking among competing goals rather than com- peting responses (Dreher et al., 2008; Shallice & Burgess, 1991).
So why do we need to perform many motor tasks slowly, step by step (for example, performing long divi- sion to find the answer to 247 divided
by 13), although we can perform many kinds of much more complex sensory tasks simultaneously, quickly, and ef- fortlessly (for example, recognizing a face as male or female)? The answer may lie in the differing architecture of motor control versus sensory per- ception systems. Sensory perception involves integrating multiple features of a stimulus into a final result—a set of computations that can be done si- multaneously (in parallel), to arrive at an answer quickly (FIGURE 7.29a). Most
FIGURE 7.28 An fMRI study of multitasking. (a) Subjects had to press buttons according to an auditory cue and then, after a short interval (SOA), pronounce a syllable according to a visual cue (or vice versa). When the intervals were short, reaction times were delayed, suggesting a “bottleneck” in information processing. (b) fMRI localized this bottleneck to the medial and lateral prefrontal cortex, which were slower to activate as reaction times increased during multitasking (Dux et al., 2006).
(b)
(a)
Auditory stimulus
Manual response
Visual stimulus
Vocal response
Visual stimulus
SOA
SOA
Vocal response
Auditory stimulus
Manual response
Supplementary motor area Posterior lateral prefrontal cortex
T1 VVAM T2 VVAM T1 AMVV T2 AMVV
R ea
ct io
n t
im e
(s ec
)
SOA
0.5
1
1.5
2
Short Long
07-Eagleman_Chap07.indd 225 09/10/15 2:28 AM
LE A R NING OBJECTI V ES provide a guide to what you will read and learn, helping you focus on the most im- portant points. Each learning objective corresponds to a major section of text.
37
# 158305 Cust: OUP Au: Eagleman Pg. No. 36 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
# 158305 Cust: OUP Au: Eagleman Pg. No. 37 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Describe the basic underlying organization of all
vertebrate central nervous systems.
• Summarize the basic organization and structure of the peripheral nervous system.
• Explain the circuitry and function of spinal reflexes and central pattern generators.
• Distinguish the major components of the brainstem and their functions.
• Characterize the anatomy of the cerebellum and its role in motor function.
• Illustrate the role of the hypothalamus in homeostasis and the role of the thalamus as a relay and synchronization center, using examples.
• Identify the locations of the four lobes of the cerebral cortex, the locations of the major gyri and sulci, and their functions.
• Characterize the components of the basal ganglia and their functions.
• Distinguish the major components of the limbic system and their functions.
36
The Brain and Nervous System
STARTING OUT: The Brains of Creatures Great and Small
An Overview of the Nervous System The Peripheral Nervous System The Spinal Cord
CASE STUDY: Christopher Reeve, 1952–2004
THE BIGGER PICTURE: In Search of a Cure for Spinal Cord Injury
The Brainstem
NEUROSCIENCE OF EVERYDAY LIFE: Why Do We Get the Hiccups?
The Cerebellum The Diencephalon: Hypothalamus and Thalamus
CASE STUDY: Waking the Brain
The Telencephalon: Cerebral Cortex and Basal Ganglia
RESEARCH METHODS: Cytoarchitecture of the Cortex
Uniting the Inside and Outside Worlds
CHAPTER 2
02-Eagleman_Chap02.indd 36-37 28/10/15 1:41 pm
STA RTING OUT scenarios begin each chapter with a gripping real-world example of chapter concepts, from hikers on Mount Everest who make a deadly mistake to a boy who functions normally following the removal of half of his brain.
104 PART 1 • ChAPTeR 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 104 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT: The Child with half a Brain By the time Matthew S. was 6 years old, he began to have several epi- leptic seizures each day—some- times every few minutes. Medications were of no use. He was diagnosed with Rasmussen’s en- cephalitis, a rare, chronic inflam- matory disease that typically affects only a single brain hemisphere. His parents explored their options and were shocked to learn that there was only one known treatment for Rasmussen’s: removal of an entire hemisphere of the brain (Borgstein & Gottendorst, 2002).
But how could it be possible to live with half of the brain missing? Aren’t the functions of the brain dis- tributed widely across its territo- ries? Wouldn’t removal of one half be fatal—or at least devastating to Matthew’s quality of life?
With no remaining options, Mat- thew’s parents took him to Johns Hopkins Hospital in Baltimore, Maryland, where he underwent a hemispherectomy: the complete removal of half the cerebrum (FIGURE 4.1). The empty half of the skull filled up with cerebrospinal fluid, which shows up as a black void in neuroimaging.
Matthew walks with a slight limp on the opposite side of his body. Otherwise, he lives a normal life with almost no measurable deficit in cognition or behavior. How
can this be possible? Because the remainder of his brain has dynami- cally rewired to take over the miss- ing functions. The normal maps of the brain have redrawn themselves
on a smaller piece of neural real estate. How the brain accomplishes this remarkable feat—something no manmade machine can yet do— is the subject of this chapter.
FIGURE 4.1 Hemispherectomy. In a hemispherectomy, half the brain is surgically removed. This surgery has become standard operating procedure for Rasmussen’s encephalitis, a rare inflammatory disease that often affects only one hemisphere. Amazingly, as long as the surgery is performed before the age of 8, the child does remarkably well: the remainder of the brain dynamically rewires to take over the missing functions.
The brain is often thought of as a fixed organ with different regions dedicated to specific tasks. But the brain is better un- derstood as a dynamic system, constantly modifying its own circuitry to match the demands of the environment and the goals of the animal. This ongoing rewiring is the brain’s most fundamental principle and the source of its utility. W hereas your computer is built with hardwiring that remains fixed
from the assembly line onward, the brain dynamically recon- figures, ever so subtly, with each new experience. It reorga- nizes itself from the level of molecules in the synapses to the level of the gross anatomy visible to the naked eye. W hen you learn something new (such as your professor’s name), your brain physically changes. This ability to physically change, and to hold that change, is known as plasticity—just like the
04-Eagleman_Chap04.indd 104 28/10/15 3:04 pm
00-Eagleman-FM.indd 16 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xvii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Preface xvii
A CLE A R, MODER N A RT PROGRA M provides at- tractive biological drawings to help convey important con- cepts and information. Photographs and historical images also connect chapter content to the world around us.
316 PART 3 • CHAPTER 10 Sleep
rhythms in the absence of light cues like the sun (Rado, Gev, Goldman, & Terkel, 1991). Incredibly, this complex network of biochemical cascades results in a clock whose natural period is 24 hours and 11 minutes. Across subjects, the vari- ation is remarkably small: plus or minus only 16 minutes (Czeisler et al., 1999).
� e primary clock in mammals lies in the suprachias- matic nucleus (SCN) of the hypothalamus (from supra meaning “above” and chiasmatic referring to the optic chiasm, or crossing of the optic nerves; the two nuclei (le� and right hemisphere) are located just above this crossing, see FIGURE 10 . 9 ) (K lein, Moore, & Reppert, 1991). Cells of the SCN maintain their own rhythm when cultured in a dish, and damage of the SCN obliterates a regular sleep–wake rhythm in animals (Ibuka & Kawamura, 1975; Welsh, Logo- thetis, Meister, & Reppert, 1995).
Entrainment of the Circadian Rhythm by Light Cues A lthough the circadian rhythm is endogenously generated, it becomes entrained to various environmental stimuli, known as zeitgebers (Ascho� , Daan, & Honma, 1982). � e most important of these is the light–dark cycle—that is, the phase of the circadian rhy thm is set by the planet’s rota- tion into and out of light from the sun. � e SCN receives information about external light levels from the eyes, but not via the rods or cones that we learned about in Chapter 5. � e retina contains a third type of light-sensitive cell: retinal
The Circadian Rhythm � ere are two important aspects to the sleep–wake cycle: how much you sleep and when you sleep. As for how much , not everyone needs to sleep the same amount. In fact, most people sleep about 6.5–7 hours a night, with a range between 4 and 11 hours (National Sleep Foundation, 2005, 2011, 2013). How many hours you sleep correlates with what you have done during the day, and periods of intense stimulation (such as going to a museum, an amusement park, and so on) during the day tend to make people sleep longer that night (Shapiro, Bortz, Mitchell, Bartel, & Jooste, 1981). How much sleep you require in general appears to have at least some ge- netic basis, although this is poorly understood (Franken, Malafosse, & Ta� i, 1999).
� e rest of this section discusses the second half of the story: when you sleep. W hy is sleep on such a regular cycle, and with what environmental factors is the sleep cycle syn- chronized? � e answer to that question involves the circa- dian rhythm , a natural internal rhythm that runs on an approx imately 24-hour cycle ( circa meaning “about” and dian referring to a “day”) and controls our sleep–wake cycles (R ichardson, 2005). A ll animals appear to have some form of circadian cycle. � is circadian rhythm in� uences not just sleep and wakefulness, but also coordination, blood pres- sure, alertness, and body temperature ( FIGURE 10 . 8 ).
� e circadian rhythm is endogenously generated , meaning that it comes from programmed mechanisms in our brain and persists even in the absence of external cues. Some animals, like the blind mole rat, maintain their endogenous
FIGURE 10 . 8 Physiological changes tied to the circadian rhy thm. The circadian rhythm runs on an approximately 24-hour c ycle and controls sleep-wake cycles. It also influences other physiological and cognitive processes, including temperature, alertness, blood pressure, and hormone levels.
14:30 Best coordination
15:30 Best reaction time
17:00 Greatest heart efficiency and muscle strength
18:30 Highest blood pressure
19:00 Highest body temperature
21:00 Melatonin secretion begins2:00
Deepest sleep
4:30 Lowest body temperature
6:45 Rise in blood pressure
7:30 Melatonin secretion stops
9:00 High testosterone secretion
10:00 High alertness
00:00 Midnight
18:006:00
12:00 Noon
Chapter Ending K EY PR INCIPLES summarize the main points covered in the chapter—with one principle corresponding to each main heading—to remind you of what you have learned.
# 158305 Cust: OUP Au: Eagleman Pg. No. 469 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Key Principles 469
Conclusion The architecture of motivation and reward is ancient and es- sential to survival. Motivations arise out of basic survival needs, and resources become rewards when they help us meet those needs. The brain has an elaborate and ancient mechanism for learning to predict the value of rewards in meeting future needs. This mechanism relies on accurate measurements of the internal state and accurate measure- ments of prediction error: the difference between the ex- pected and the actual outcomes. W hen these mechanisms operate normally, they are effective at keeping us alive.
Unfortunately, these mechanisms can be hijacked, with devastating consequences. By altering neurotransmitter sig- nals, some substances can create powerful illusions of well- being and illusions of always being better than expected. W hen this happens, a vicious circle of pathological learning begins. This pathological learning creates an unnaturally strong motivation to obtain and consume the addictive sub- stance. The strength of this motivation increases over subse- quent exposures, in time eclipsing other survival needs.
Undoing this process is as difficult as unlearning a skill. Current treatments for addiction rely on a combination of counseling and, in some cases, medications to alleviate with- drawal and reduce the high risk of relapse. Unfortunately, effective treatments for addiction have been hard to find, and substance abuse ranks among the most devastating causes of disability and death worldwide. Finding a safe and effective treatment for substance dependence is one of the major goals in the neuroscience of the 21st century.
Paradox ically, this often requires that the doctor take the part of discussing the reasons not to change. The pa- tient then has no option but to take up the other side of the argument. For example, a physician might begin by ask ing the patient, “On a scale of 1 to 10, how much do you want to quit smok ing?” The patient might reply, “Not much, maybe a 2.” A lthough the physician’s natural reaction might be to splutter, “ Don’t you realize the damage that smok ing does to your health?”, such an approach only makes the patient defensive, encouraging him to generate self- conv incing reasons why it’s too hard or too stressf ul to quit right now. A doctor trained in MI might instead say, “A 2 out of 10 . . . OK, so why not a 1? W hy not a zero?” This gets the person think ing and talk ing about the reasons they actually do want to change: setting a good example for their children, not wanting to leave their spouse alone and bereft in old age. By eliciting the other side of the argu- ment, the physician gets the patient to think a little harder about his or her ow n motivations for overcoming the habit of smok ing. With time and discussion, these motivations move to the forefront, increasing the chances that the person w ill make a break from prior habits (Rollnick, Miller, & Butler, 2007).
MI is not a cure- all, but it seems to work better than be- rating people for their bad habits in the hopes this will per- suade them to change. Its take- home point: the motivation to change is hard to implant in someone from the outside. Instead, it must be found within the person’s own set of values, goals, hopes, and ideals and then carefully coaxed out into the forefront. So next time you find yourself “wanting someone to want” something, give this approach a try.
KEY PRINCIPLES
• Natural motivations, such as eating, drinking, and reproductive behaviors, help inform the brain what is needed and how to value those needs at the cur- rent time.
• The hypothalamus is important for homeostasis and for evaluating internally driven motivation, whereas the amygdala receives input from the ex- ternal world and evaluates the importance of these outside factors. Dopamine is the neurotransmitter that most commonly conveys these reward and motivation signals.
• The brain learns to predict rewards by comparing the expected outcome of an action to the actual outcome. “ Better- than-expected” outcomes in- crease the motivation toward that action in the future.
• “Liking” and “wanting” are two different things in the brain. Liking refers to the sensation of well- being that is being experienced at that time, whereas wanting refers to a future expectation of well- being.
• Opioids are naturally occurring chemicals in the brain that reduce pain and increase pleasure. These effects can be mimicked by synthetic opi- oids, such as morphine, heroin, and codeine. Recent research has identified several types of opioid receptors. Stimulation of some of these re- duces pain, but stimulation of others produces un- pleasant sensations.
• The neurotransmitter dopamine is important for motivation and in learning to predict rewards. It is especially important for assigning value to those
14-Eagleman-Chap14.indd 469 28/10/15 3:09 pm K EY TER MS include all of the chapter’s bold glossary
terms listed by subsection, with page numbers, for easy exam review.
THE BIGGER PICTUR E connects neuroscience to larger concepts and questions—social, ethical, legal, and historical. Examples include how neuroscience can help us make better decisions and whether we will one day be able to equip our brains with new senses.
214 PART 2 • ChAPTeR 7 The Motor System
# 158305 Cust: OUP Au: Eagleman Pg. No. 214 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
THE BIGGER PICTURE Neural Implants for Motor Control Remember the tragic case of Jean- Dominique Bauby from the begin- ning of the chapter? He had suffered a stroke affecting the medulla and the nearby pyramidal decussations: the outgoing pathway for virtually all of the motor cortex. Locked-in syndrome is the usual result of this kind of stroke. Without any connec- tions to the spinal cord or lower cra- nial nerves, the cortex is unable to send output to nearly any parts of the body. A few cranial nerves above the injury may be spared, allowing some movement of the eyes and eyelids only. Since these pathways do not easily regenerate, the hopes for recovery are slim.
Yet the motor cor tex itself is still intact. What if there were some way to read the activity of the upper motor neurons directly? Could we use them to drive an ar- tificial arm, or a wheelchair, or a computer cursor? Perhaps even a simple speech synthesizer? In fact, for more than a decade, neu- roscientists have been using brain– computer interface tech- nologies to help locked-in patients communicate with the outside world (Kennedy, Bakay, Moore, Adams, & Gold waithe, 2000). One strategy involves implanting a set of electrodes directly in the motor cor tex, in the par t of the homuncu- lus that controls the hand or the mouth. The electrodes may be coated with neural growth factors to encourage the neurons to grow connections to the implant itself.
As one example, in 2006, a re- search team from Massachusetts General Hospital and Brown Univer- sity inserted a small electrode array (FIGURE 7.17) into the motor cortex of Matthew N., a 25-year-old who had
been quadriplegic since a knife attack five years earlier. A success- ful example of “neuromotor pros- thetics” in humans, the sensor was able to read the signals that Mat- thew’s brain was trying to send out to his body, and convert those sig- nals into commands that could direct the movements of a pros- thetic arm or a pointer on a com- puter screen (Hochberg et al., 2006).
Decoding the neural activity into meaningful signals can be difficult. Computer algorithms can learn to in- terpret the collective activity of a pop- ulation of neurons as signaling a particular movement, or syllable. The patients themselves can learn to use the prostheses, rewiring the local connections to improve their ability to communicate over time. More
recently, a young man with locked-in syndrome has begun to use this system to produce simple syllables with reasonable accuracy after sev- eral years of training for both patient and computer (Guenther et al., 2009).
So far, the technology is still in its infancy. Interpreting the neural signals remains difficult. The sig- nals themselves fade over time as glial cells gradually build up around the implanted electrodes, render- ing the electrodes useless. At pres- ent, no implant has been able to match the efficacy of the simple blink-coding card used by M. Bauby to write his memoirs. Developing effective brain–computer inter- faces will be one of the most impor- tant technological challenges of the 21st century.
FIGURE 7.17 A neuromotor prosthetic. A tiny electrode array implanted in the motor cortex of Matthew N. enabled him to control his wheelchair and actions on a screen.
07-Eagleman_Chap07.indd 214 28/10/15 3:06 pmR ESE A RCH M ETHODS shows how we know what we know about the brain, presenting important research tech- niques and indicating the types of research questions that these techniques have been used to investigate.
The Cells of the Brain 81
# 158305 Cust: OUP Au: Eagleman Pg. No. 81 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
RESEARCH METHODS: Visualizing Neurons and Their Products The discovery that neurons are dis- crete, fundamental units of the ner- vous system was made possible by the ability to stain them. Several techniques allow the visualization of neurons. Golgi staining is a tech- nique that impregnates some frac- tion of neurons with a dark material, allowing the entirety of individual cells to be seen under a microscope (FIGURE 3.10a). This is the method that birthed Ramón y Cajal’s neuron doctrine. Another technique, Nissl staining, uses a chemical that binds to the RNA in cell bodies, thereby allowing the visualization of somas (FIGURE 3.10b). Nissl staining is most commonly used for judging sizes of cells and their densities.
Several other methods are utilized to obtain detailed pictures of nervous tissue. In the technique of autoradi- ography, a radioactive substance is designed to be taken up by specific cells but not by others (FIGURE 3.10c). Then, when a photographic emulsion is placed over thin slices of the brain tissue, the emulsion is exposed by the radioactivity in the same way that film is exposed by light. In this way, it can be seen which cell types absorbed the substance in question (for example, a pharmaceutical drug).
In the technique of immunocyto- chemistry, antibodies are developed that bind only to specific proteins (FIGURE 3.10d). These antibodies are washed onto a slice of brain tissue, and they attach wherever the protein of interest is being expressed. With some chemical steps, these antibod- ies can be visualized, revealing the
exact locations of the protein within the cell. A related technique is to use radioactively labeled stretches of RNA or DNA that will bind to specific
stretches of messenger RNA (mRNA); this is called in situ hybridization, and it reveals which cells have expressed a gene of interest (FIGURE 3.10e).
A
C
FIGURE 3.10 Different techniques to bring the invisibly small world of neurons to light. (a) Golgi staining, (b) Nissl staining, (c) autoradiography, (d) immunocytochemistry, and (e) in situ hybridization.
(a) (b)
(d)
(e)(c)
03-Eagleman_Chap03.indd 81 09/10/15 2:30 AM
00-Eagleman-FM.indd 17 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xviii Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
xviii Preface
• Chapter Summaries: Full summaries of each chapter provide a thorough review of the important facts and con- cepts covered.
• Flashcards: Interactive flashcard activities are an effec- tive way for students to learn and review all of the impor- tant terminology.
• Practice Quizzes: Each chapter includes a practice quiz, which students can use as a self-review exercise, to check their understanding.
For Instructors An extensive and thoughtful supplements program offers in- structors everything they need to prepare their course and lectures, and assess student progress.
Ancillary Resource Center (ARC) For more information, go to w w w.oup.com/us/eagleman
Available online exclusively to adopters, the Ancillary Resource Center (A RC) includes all of the instructor re- sources that accompany Brain and Behavior: A Cognitive Neuroscience Perspective.
Instructor’s Manual: For each chapter of the textbook, the Instructor’s Manual includes the following: • Chapter Overview • Chapter Outlines • Key Concepts • Suggested Online Activities • Journal Articles and Press Releases
Textbook Figures and Tables: A ll of the textbook ’s il- lustrations and tables are provided in a variety of formats, including high and low resolution, with and without balloon captions, and unlabeled (all balloon captions, labels, and leaders removed).
PowerPoint Resources: • Figures and Tables: This presentation includes all of the
figures and tables (all formats) from the chapter, with titles.
• Lecture: A complete lecture outline, ready for use in class. Includes coverage of all important facts and concepts pre- sented in the chapter along with selected figures and tables. Animations: A ll of the animations from Dashboard are
available in the A RC for download, making it easy to include them in lecture presentation and online course materials. (A lso available in Dashboard.)
Videos: A collection of videos selected to accompany each chapter helps bring some of the key concepts from the textbook to life. Ideal for use as lecture starters or paired with assignments.
Test Bank: A complete test bank provides instructors with a wide range of test items for each chapter, including multiple-choice, fill-in-the-blank, short-answer, true/false, and essay questions. Questions are noted for whether they
R EV IEW QUESTIONS test your recall and under- standing of the key information presented in the chapter to aid with studying.
# 158305 Cust: OUP Au: Eagleman Pg. No. 555 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical- Thinking Questions 555
REVIEW QUESTIONS
1. In what parts of the brain do we tend to see the plaques and tangles of Alzheimer’s disease? Why do you think that removing these plaques and tangles turned out to be a failure as a strat- egy for treating Alzheimer’s disease?
2. Is it a reasonable metaphor to call frontotempo- ral dementia a “cancer of the soul?” What does this name tell us about our own intuitions on which faculties are proper to the “spirit,” as op- posed to the neural circuits of the brain?
3. Why does increased creativity emerge as a symptom of some neurological and psychiatric disorders? What do these disorders have in common?
4. In Huntington’s disease, a disruption of a single gene gives rise to a diverse set of neurological and psychiatric symptoms. What are these symptoms, what common disease process cre- ates them, and how does this common process disrupt such a wide variety of brain functions?
5. What different kinds of symptoms can arise in Tourette syndrome? What common neuro- anatomical changes might give rise to these symptoms?
6. Should we describe symptoms of Tourette syn- drome as “voluntary” or “involuntary”? Why so? Should ordinary blinking be considered a kind of motor tic? Why or why not?
7. Obsessive–compulsive disorder is traditionally considered a psychiatric rather than a neurolog- ical disorder. How would you classify the illness? Why? Do you think it should still be considered a single category of disease? Why or why not?
8. What are the positive and the negative symp- toms of schizophrenia? Which symptoms can be relieved with currently available treatments and which cannot? Why don’t these treatments provide a permanent cure?
9. What are three kinds of similarities between bi- polar disorder and schizophrenia? What are some of the key differences between these dis- orders? Should we think of them as two forms of the same illness? Why or why not?
10. What kinds of factors can lead to depression? Is depression really a chemical imbalance in the brain, as it is sometimes described? What and where are the brain abnormalities of depression? How has this knowledge led to new treatments?
CRITICAL- THINKING QUESTIONS
1. Imagine that you are the head of a pharmaceuti- cal research group that is dedicated to develop- ing new, effective drug treatments for Alzheimer’s disease. Given what you have learned in this chapter about the physiological basis and progression of Alzheimer’s, what pro- cesses would you design the drugs to target? At what stage of the disease do you think that such drugs would be most effective? Explain.
2. How do patients with Alzheimer’s disease, fron- totemporal dementia, Tourette syndrome, obsessive–compulsive disorder, and bipolar disorder differ from one another in terms of their self- awareness of their symptoms? How
could such differences be used to guide the de- velopment of new, specialized treatments for each disease?
3. Scientists are continuing to learn more about the influence of genetics on the likelihood of de- veloping diseases such as schizophrenia. How- ever, we are unable to know for certain whether individuals are destined to develop such dis- eases. If an effective neuroprotective treatment were discovered for schizophrenia, do you think it would be ethical to administer it to individuals who are at risk for developing the disease? What factors would influence your viewpoint? Explain your reasoning.
16-Eagleman-ch16.indd 555 08-Jul-15 8:19:12 PM
CR ITICA L-THINK ING QUESTIONS ask you to apply and extend information from the chapter to new sce- narios to help you master the material.
# 158305 Cust: OUP Au: Eagleman Pg. No. 555 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical- Thinking Questions 555
REVIEW QUESTIONS
1. In what parts of the brain do we tend to see the plaques and tangles of Alzheimer’s disease? Why do you think that removing these plaques and tangles turned out to be a failure as a strat- egy for treating Alzheimer’s disease?
2. Is it a reasonable metaphor to call frontotempo- ral dementia a “cancer of the soul?” What does this name tell us about our own intuitions on which faculties are proper to the “spirit,” as op- posed to the neural circuits of the brain?
3. Why does increased creativity emerge as a symptom of some neurological and psychiatric disorders? What do these disorders have in common?
4. In Huntington’s disease, a disruption of a single gene gives rise to a diverse set of neurological and psychiatric symptoms. What are these symptoms, what common disease process cre- ates them, and how does this common process disrupt such a wide variety of brain functions?
5. What different kinds of symptoms can arise in Tourette syndrome? What common neuro- anatomical changes might give rise to these symptoms?
6. Should we describe symptoms of Tourette syn- drome as “voluntary” or “involuntary”? Why so? Should ordinary blinking be considered a kind of motor tic? Why or why not?
7. Obsessive–compulsive disorder is traditionally considered a psychiatric rather than a neurolog- ical disorder. How would you classify the illness? Why? Do you think it should still be considered a single category of disease? Why or why not?
8. What are the positive and the negative symp- toms of schizophrenia? Which symptoms can be relieved with currently available treatments and which cannot? Why don’t these treatments provide a permanent cure?
9. What are three kinds of similarities between bi- polar disorder and schizophrenia? What are some of the key differences between these dis- orders? Should we think of them as two forms of the same illness? Why or why not?
10. What kinds of factors can lead to depression? Is depression really a chemical imbalance in the brain, as it is sometimes described? What and where are the brain abnormalities of depression? How has this knowledge led to new treatments?
CRITICAL- THINKING QUESTIONS
1. Imagine that you are the head of a pharmaceuti- cal research group that is dedicated to develop- ing new, effective drug treatments for Alzheimer’s disease. Given what you have learned in this chapter about the physiological basis and progression of Alzheimer’s, what pro- cesses would you design the drugs to target? At what stage of the disease do you think that such drugs would be most effective? Explain.
2. How do patients with Alzheimer’s disease, fron- totemporal dementia, Tourette syndrome, obsessive–compulsive disorder, and bipolar disorder differ from one another in terms of their self- awareness of their symptoms? How
could such differences be used to guide the de- velopment of new, specialized treatments for each disease?
3. Scientists are continuing to learn more about the influence of genetics on the likelihood of de- veloping diseases such as schizophrenia. How- ever, we are unable to know for certain whether individuals are destined to develop such dis- eases. If an effective neuroprotective treatment were discovered for schizophrenia, do you think it would be ethical to administer it to individuals who are at risk for developing the disease? What factors would influence your viewpoint? Explain your reasoning.
16-Eagleman-ch16.indd 555 08-Jul-15 8:19:12 PM
Media and Supplements to Accompany Brain and Behavior: A Cognitive Neuroscience Perspective
For Students
Companion Website Available at no additional cost, the Companion Website pro- vides students with the following review resources: • Chapter Outlines: Detailed outlines give an overview of
each chapter.
00-Eagleman-FM.indd 18 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xix Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Preface xix
progress instantly. Dashboard includes the follow ing resources: • Quizzes: For each chapter of the textbook, there is a quiz
to test student comprehension of important facts and concepts introduced in the chapter.
• A postlecture summative quiz designed to be used as an assessment of student mastery of the important facts and concepts introduced in the chapter, after the student has read the chapter and attended the relevant lecture/class period/discussion section.
• Animations: A set of detailed animations helps students understand some of the book ’s more complex topics and processes by presenting them in a clear, easy-to-follow narrative.
LMS Course Cartridges For those instructors who wish to use their campus learning management system, a course cartridge containing all of the Dashboard resources is available for a variety of e-learning environments. (For more information, please contact your local Oxford representative.)
are factual or conceptual, and for level of difficulty. A ll ques- tions from the Dashboard and Companion Website quizzes (see below) are also included.
Computerized Test Bank: The Test Bank is also pro- vided in Blackboard Diploma format (software included). Diploma makes it easy to create quizzes and exams using any combination of publisher-provided questions and an in- structor’s own questions and to export those assessments for print or online delivery in a wide range of learning manage- ment system formats.
Dashboard For more information, go to w w w.oup.com/us/dashboard
Oxford’s Dashboard learning management system fea- tures a streamlined interface that connects instructors and students with the functions they perform most often, simpli- f y ing the learning ex perience to save instr uctors time and put students’ progress f irst. Dashboard ’s prebuilt assess- ments were created specif ically to accompany Brain and Behavior: A Cognitive Neuroscience Perspective and are au- tomatically graded so that instr uctors can see student
Acknowledgments Over the past several years, many talented brains have devoted their cognitive powers to making Brain and Behavior a reality. We express our heartfelt gratitude for the hard work, patience, and dedication of the team at Oxford University Press, without whom this book would not exist. Special thanks go to editorial director Patrick Lynch, who first saw the project’s potential, as well as John Challice, vice president and publisher, who also supported the book at an early stage. Kind thanks to our editor, Jane Potter, who encouraged us steadily onward from deadline to deadline, with patience and gentle persuasion. We are also ever grateful to development editor Anne Kemper, senior devel- opment editor Lisa Sussman, and assistant editor Maura Mac- Donald for their careful attention to detail throughout the manuscript. We also thank the Oxford production team for transforming the unadorned drafts of the manuscript chapters into such an eye-catching and engaging final form: Lisa Grzan,
production manager, Jane Lee and Keith Faivre, senior produc- tion editors, and Susan Brown, copyeditor. Kudos to art direc- tor Michele Laseau and senior designer Caitlin Wagner for their truly beautiful work on the interior and the cover design. Thanks also for the essential efforts of Eden Gingold, marketing manager, Kateri Woody, marketing associate, and Frank Mor- timer, director of marketing, for helping to bring this book before a wide audience of curious minds. Grateful acknowledg- ment is made to the talents of the team at Dragonfly Media for the art program for the book—specifically, art development and art direction by Mike Demaray; art production by Mike Demaray, Craig Durant, Helen Wortham, and Rob Fedirko; and chapter openers and cover art by Craig Durant. Finally, thanks to all of our reviewers, anonymous and otherwise, for their suggestions and insightful comments on the early drafts of this work:
Lewis Barker, Auburn University
Diane E. Beals, University of Tulsa
Patricia Bellas, Irvine Valley College
Annemarie Bettica, Manhattanville College
Christopher Braun, Hunter College
Blaine Browne, Valdosta State University
David Bucci, Dartmouth College
Amanda N. Carey, Simmons College
Cynthia R . Cimino, University of South Florida
Barbara Clancy, University of Central Arkansas
Howard Casey Cromwell, Bowling Green State University
00-Eagleman-FM.indd 19 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xx Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
xx Preface
Kelly L. Curtis, High Point University
Deana Davalos, Colorado State University
Scott Decker, University of South Carolina
Dean Dessem, University of Maryland
Vonetta Dotson, University of Florida
Jeffrey Eells, Mississippi State University–Main
Paul Engelhardt, Michigan State University
Joseph Farley, Indiana University–Bloomington
Robert Faux, Duquesne University
Robert P. Ferguson, Buena Vista University
Jane Flinn, George Mason University
Jay Friedenberg, Manhattan College
Jonathan Gewirtz, University of Minnesota
Edward Golob, Tulane University
K im Gorgens, University of Denver
Jinger S. Gottschall, Penn State University
Jay E. Gould, University of West Florida
Sayamwong E. Hammack, University of Vermont
Valerie Gray Hardcastle, University of Cincinnati
Linda Hermer, University of Florida
Elaine M. Hull, Florida State University
Daniel Hummer, Morehouse College
Mark Hurd, University of Texas
Eric Jackson, University of New Mexico
Daniel Jacobson, Madonna University
Mark Jareb, Sacred Heart University
Penelope L. Kuhn, California State University–Chico
Matthew Kurtz, Wesleyan University
Eric Laws, Longwood University
Ben Lester, University of Iowa
Linda Lockwood, Metropolitan State College of Denver
Jeannie Loeb, University of North Carolina–Chapel Hill
Keith B. Lyle, University of Louisville
Cyrille Magne, Middle Tennessee State University
Kai McCormack, Spelman College
Ming Meng, Dartmouth College
Maura Mitrushina, California State Northridge
Daniel Montoya, Fayetteville State University
Andrea Morris, University of California–Los Angeles
Ezequiel Morsella, San Francisco State University
Andrea Nicholas, University of California–Irvine
J. Ian Norris, Berea College
Jamie Olavarria, University of Washington
Matthew Palmatier, East Tennessee State University
Jim H. Patton, Baylor University
Tadd B. Patton, Georgia Regents University
R ichard Payne, University of Maryland–College Park
Michael Sakuma, Dowling College
Haline Schendan, Plymouth University
Lynda Sharrett-Field, University of Kentucky
Robert W. Sikes, Northeastern University
Scott Slotnick, Boston College
Kenith V. Sobel, University of Central Arkansas
Jessica Stephens, Texas A&M University at K ingsville
Jeffrey Taube, Dartmouth College
Sheralee Tershner, Western New England University
Jason Themanson, Illinois Wesleyan University
Lucien T. Thompson, University of Texas at Dallas
Lucy J. Troup, Colorado State University
Jonathan Vaughan, Hamilton College
Sandy Venneman, University of Houston–Victoria
Todd D. Watson, Lewis & Clark College
Douglas A. Weldon, Hamilton College
Robin Wellington, St. John's University
Mark West, Rutgers University
Adrienne Williamson, Kennesaw State University
John L. Woodard, Wayne State University
With special thanks to our Advisory Panel:
A lex Michael Babcock, Montana State University
Peter Brunjes, University of Virginia
Arne Ekstrom, University of California–Davis
Samuel McClure, Stanford University
Juan Salinas, University of Texas at Austin
00-Eagleman-FM.indd 20 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. xxi Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
David Eagleman is a neuroscientist, New York Times best- selling author, and Guggenheim Fellow who holds joint ap- pointments in the departments of neuroscience and psychiatry at Baylor College of Medicine in Houston, Texas. Dr. Eagleman’s areas of research include time perception, vision, synesthesia, and the intersection of neuroscience with the legal system. He directs the Laboratory for Percep- tion and Action and is the founder and director of Baylor College of Medicine’s Initiative on Neuroscience and Law. Dr. Eagleman has written several neuroscience books, in- cluding Incognito: The Secret Lives of the Brain and Wednesday Is Indigo Blue: Discovering the Brain of Synesthesia. He has also written an internationally bestselling book of literary fiction, Sum, which has been translated into 28 languages and turned into two operas in Sydney and London. Dr. Eagleman is the author and presenter of “The Brain,” an international six- hour series on PBS that poses the question, “W hat does it mean to be human?” from a neuroscientist’s point of view. Dr. Eagleman has written for the Atlantic, New York Times, Discover Magazine, Slate, Wired, and New Scientist and ap- pears regularly on National Public Radio and BBC.
Jonathan Downar is the Director of the MR I-Guided rTMS Clinic at the University Health Network in Toronto, Canada, and a scientist at the Toronto Western Research Institute. He currently holds appointments with the Department of Psychiatry and the Institute of Medical Science at the Uni- versity of Toronto.
As a physician-scientist, his clinical work focuses on using noninvasive brain stimulation to treat patients with severe and medication-resistant forms of psychiatric illness, including depression, bipolar disorder, obsessive– compulsive disorder, post–traumatic stress disorder, and eating disor- ders. His research work focuses on developing a new genera- tion of more effective, more accessible, and less costly techniques for brain stimulation in these disorders. His re- search laboratory also focuses on developing tests that use functional MR I and EEG to predict the most effective treat- ment parameters for individual patients.
In addition to his research and clinical work, he teaches undergraduate courses in the neuroscience of social cogni- tion, emotion regulation, decision making, and other forms of complex human behavior. He also teaches medical students and psychiatry resident physicians on the subjects of neuro- anatomy, neuroimaging, and therapeutic brain stimulation.
ABOUT THE AUTHORS
00-Eagleman-FM.indd 21 29/10/15 2:55 pm
00-Eagleman-FM.indd 22 29/10/15 2:55 pm
A Cognitive Neuroscience Perspective
Brain and Behavior
00-Eagleman-FM.indd 1 29/10/15 2:55 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 2 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
2
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Describe the central aims of cognitive
neuroscience.
• Distinguish between understanding the brain’s components and understanding its functioning.
• Compare and contrast the advantages and disadvantages of each of four categories of research methods for studying brain function.
• Summarize three common forms of cognitive bias in human thinking and how the key elements of the scientific method address cognitive biases.
• Describe five big questions in cognitive neuroscience.
• Show at least three ways in which advances in neuroscience may benefit human society.
01-Eagleman_Chap01.indd 2 02/11/15 3:08 pm
3
# 158305 Cust: OUP Au: Eagleman Pg. No. 3 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Introduction STARTING OUT: A Spark of Awe in the Darkness
Who Are We? In Pursuit of Principles How We Know What We Know
RESEARCH METHODS: Magnetic Resonance Imaging
Thinking Critically about the Brain The Big Questions in Cognitive Neuroscience The Payoffs of Cognitive Neuroscience
CHAPTER 1
01-Eagleman_Chap01.indd 3 02/11/15 3:08 pm
4 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 4 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT: A Spark of Awe in the Darkness On October 9, 1604, a brilliant spark of light grew to life in the darkness of the night sky over Europe. A few days later, the astronomer Johannes Kepler began to gaze up at the new star that had appeared in the void, outshining all its peers, visible for a time even through the brightness of the day. Kepler wrote extensively on the astronomical properties of the new star, or stella nova, whose sudden appearance challenged the conventional wisdom that the heavens were fixed and unchang- ing (Kepler, [1604] 2004). Over the ensuing months, the new star faded gradually back into the ce- lestial background. Nothing simi- lar has appeared in our skies to surpass it since then, even four centuries later.
Today’s astronomers would have called Kepler’s star a super- nova and could have told him some astonishing details about the nature of the object that captured his attention on that clear night so long ago (FIGURE 1.1). They could have told him about a star several times more massive than the Sun, reaching the end of a lifespan mea- sured in eons, collapsing suddenly in upon itself to form a core blazing at a hundred billion degrees, then bursting outward again in a cata- clysmic explosion that, for a time, shone brighter than the entire sur- rounding galaxy. The light of that distant explosion was obliged to sear through space to arrive tens of thousands of years later in the night sky over Europe, drawing human eyes upward in wonder.
Yet a supernova, for all its mag- nificence and rarity, is still outshined
by marvels closer to home. As Kepler stood under the stars, be- holding the bright spark in the dark- ness before his eyes, even rarer and more wondrous events were taking place in the darkness behind his eyes. In that mysterious vault within his skull, a spark, not of light but of awe, was taking form in the warm, dark passageways of his brain. Where the supernova burned with common light and heat, his mind burned with a rare and incandes- cent emotion. Where the supernova shed its light blindly in all directions, his mind turned its attention to one tiny facet of the universe beyond, striving to see more clearly. Where the supernova was unaware of its own grandeur, Kepler’s mind was capable of reflecting on the curious mystery of its own existence.
Take a three-pound piece of uni- verse, arrange its atoms just so, into the knotty network of a human brain, and the resulting object de- velops remarkable properties. It is capable of knowing that it is a piece of the universe. It is capable of knowing of its own existence, ca- pable of perceiving impressions of the other bits of universe around it, and capable of thrumming with in- ternal feelings of awe, fear, joy, hatred, perplexity, and wonder. There is nothing else we know of, anywhere, that can do these things. Minds are inimitable, mysterious, and precious beyond measure. The very least of us, no matter what our failings or faults, by mere dint of being alive and aware, is more re- markable than any orb in the sky. This is what a mind is worth.
FIGURE 1.1 SN 1604, also known as Kepler’s Supernova, as seen through NASA’s Chandra X-ray Observatory in 2013. This massive stellar explosion was originally observed in 1604 by the astronomer Johannes Kepler. Supernovae of this type are rare and magnificent events: no similar explosion has been seen in our galaxy since Kepler’s time, more than 400 years ago. Yet the human brain, a three-pound piece of universe capable of thought, perception, and feeling, is arguably more remarkable than anything it might chance to observe among the stars of the night sky.
01-Eagleman_Chap01.indd 4 02/11/15 3:08 pm
Who Are We? 5
# 158305 Cust: OUP Au: Eagleman Pg. No. 5 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
began to control its own peripheral devices, removed its own cover, and pointed its webcam at its own circuitry. That’s us.
And what we’ve discovered by peering into the skull ranks among the most significant intellectual developments of our species: the recognition that the innumerable facets of our behavior, thoughts, and experience are inseparably yoked to a vast, wet, chemical–electrical network called the nervous system (FIGURE 1.3). The machinery is utterly alien to us, and yet, somehow, it is us.
The Mission of Cognitive Neuroscience The understanding of the three-pound human brain goes beyond a mere academic interest. W ho are we if not our thoughts, decisions, sensations, hopes, dreams, fears, and as- pirations? And what are these things if not the products of our brains? The field of cognitive neuroscience seeks to de- termine how the brain processes information, builds memo- ries, navigates decisions, and ultimately produces a human being from trillions of smaller parts.
How are intelligent systems built from simple, senseless parts? How is a great orator constructed from speechless cells? How is a great football player guided in direction by billions of neurons—which by themselves don’t know the rules of the game, know what a football is, or understand the concept of winning?
A lthough at first blush it seems impossible to build an acting, sentient being from neutral, ignorant parts, groups of interacting simple parts can lead to complex emergent prop- erties—that is, characteristics of a system that do not belong to any individual component. If you were to decompose your television set into its constituent resistors, capacitors, and transistors, you would see that the comedies and tragedies played out on its screen are not a property of any given piece, but of the system as a whole. The same applies to biology: if
Who Are We? Take a close look at yourself in the mirror. Beneath your dashing good looks churns a hidden universe of networked machinery. The machinery includes a sophisticated scaffold- ing of interlocking bones, a netting of sinew y muscles, a great deal of specialized fluid, and a collaboration of internal organs chugging away in darkness to keep you alive. A sheet of high-tech self-healing sensory material that we call skin seamlessly covers your machinery in a pleasing package.
And then there’s your brain: three pounds of the most complex material we’ve discovered in the universe. This is the mission control center that drives the whole operation, gathering dispatches through small portals in the armored bunker of the skull.
Your brain is built of cells called neurons and glia— hundreds of billions of them (FIGURE 1.2). Each one of these cells is as complicated as a major city. And each one contains the entire human genome and traffics billions of molecules in intricate economies. Each cell sends electrical pulses to other cells, up to hundreds of times per second. If you represented each of these trillions and trillions of pulses in your brain by a single spark of light, the combined output would be blinding.
The cells are connected to one another in a network of such staggering complexity that it bankrupts human language. A typical neuron makes about 10,000 connections to neighbor- ing neurons. Given the billions of neurons, this means there are as many connections in a single cubic centimeter of brain tissue as there are stars in the Milky Way galaxy (Nash, 1997).
The three-pound organ in your skull—with its pink con- sistency of Jell-O—is an alien kind of computational material. It is composed of miniaturized, self-configuring parts, and it vastly outstrips anything we’ve dreamt of building. So if you ever feel lazy or dull, take heart: you’re the busiest, brightest thing on the planet.
Ours is an incredible story. As far as anyone can tell, we’re the only system on the planet so complex that we’ve thrown ourselves headlong into the game of deciphering our own programming language. Imagine that your computer
FIGURE 1.2 A view of the hippocampal neurons of the human brain, visualized using confocal microscopy and fluorescent protein labeling.
FIGURE 1.3 BigBrain ultra-high-resolution three-dimensional human brain atlas, reconstructed from 7,404 stained, microscopic sections of a single human brain at a resolution of 20 micrometers.
01-Eagleman_Chap01.indd 5 02/11/15 3:08 pm
6 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 6 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
disciplines as diverse as physics, computer science, biology, psychology, philosophy, mathematics, and engineering. It is now a maturing discipline, and although it continues to draw professionals from those various disciplines, many of them consider themselves “neuroscientists” (FIGURE 1.4). The recent, expansive growth in the field also owes a debt to: (1) unprece- dented amounts of detailed biological data, (2) high-speed, low-cost computing power, and (3) rapidly maturing theoreti- cal approaches that have allowed us to see the underpinnings of many frameworks for perception, learning, reasoning, deci- sion making, and disease states.
In Pursuit of Principles
The Functions behind the Form A lthough human brains perform feats that seem almost magical, we are, after all, made of biological parts. Somehow these parts run programs that throw balls, walk along uneven paths, detect danger, lift a cup to our lips, phrase a question, communicate with facial expressions, write this sentence, read this sentence, and effortlessly perform the further pro- fusion of sophisticated activities we enjoy each hour. Cogni- tive neuroscience seeks to determine how the organization and function of the brain’s parts engender these everyday, seemingly effortless feats.
The goal is to look under the panels of the machinery of our everyday actions and behaviors to see what’s making the engine run. The challenge is that we find a universe of
you divided your body into piles of all the different molecules and cells that make you up, you would have an ensemble of uninteresting (and insentient) chemical piles. But rearrang- ing those chemicals into a particular organization, with par- ticular relationships among the molecules, can restore the motivated, dreaming, volitional creature that your friends know and love. The brain’s organization, function, and emer- gent properties are what neurobiology seeks to understand.
To comprehend the connection between mind and brain, it is necessary to ground ourselves in two concrete bodies of data: (1) the way humans behave, perceive, and decide and (2) the biological mechanisms that underlie those behaviors. In this book we will begin at the molecular level and work our way up to larger scales, highlighting the fundamental principles at work at all stages.
From the whirlwind of the brain’s great complexity we will glimpse the remarkable mystery that sensations, percep- tions, selves, minds, and even consciousness are biological products. A fter reading this book, it will be easy to under- stand how, for example, an idea in your head can cause pas- sion or peacef ulness, elevate blood pressure and pulse frequency, dispatch a hot upsurge of adrenaline, or make you feel cold with fright. Ideas lead to physical and chemical changes because they themselves are constructed of physical and chemical changes. Neural computations are required at all levels for you to read this page, just as they were required to compose it. Ideas are not incorporeal: they are con- structed of parts that are increasingly amenable to descrip- tion, and that is the journey we will take in this book.
Neuroscience Is a Relatively New Field How can biological tissue act like a special kind of computer? How do we connect neural function to cognitive capacities? W hat is the relationship between the brain and the mind, be- tween biology and psychology?
We live in a time of rapid progress in many scientific fields. Neuroscience is no exception. The word neuroscience only entered the English lexicon in the 1960s, representing a new understanding that the study of the brain—and the mind—encompasses a field of study in its own right (Schmitt, 1966). Each year remarkable progress is made in understand- ing the biology of the brain, and much of this progress centers on the detailed biophysical and biochemical processes that attend the operation of the cells comprising the brain. This progress has shone light on the physical mechanisms that permit neurons to organize and operate. The interactions of these neurons give rise to cognitive processes such as atten- tion and memory. These same approaches have yielded im- portant insights into the disease processes that attack the nervous system. These advances, when combined with the revolution in molecular biology, represent a burgeoning pic- ture of the brain and its operations.
Cognitive neuroscience has not always had a distinct name and was once the remote province of visionary thinkers from
FIGURE 1.4 Every year, nearly 30,000 researchers, fellows, students, and health professionals attend the annual meeting of the Society for Neuroscience. Their areas of expertise are diverse and include applied mathematics, biophysics, molecular and cellular biology, neuroimaging, electrophysiology, pharmacology, neurology, psychiatry, psychology, and cognitive science.
01-Eagleman_Chap01.indd 6 02/11/15 3:08 pm
In Pursuit of Principles 7
# 158305 Cust: OUP Au: Eagleman Pg. No. 7 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
characters, why they are sometimes funny and sometimes not, your reaction to their facial expressions, humorous situations, and so on. Through careful study, you may generate a theory of comedy and rules for why it has such powerful effects on humans and not on monkeys or cats. But even at the end of that theoretical work, you are still stuck with a question: ex- actly what is the relationship between the function of a tran- sistor and the comedy that is displayed on the screen?
You might turn to studying the tiny transistors and the way they are fashioned out of semiconductor materials. This would result in a description of how electrical currents and voltages relate to the detailed structure of the transistor; however, even your complete description of the transistor would give you no insight into why a video was funny.
In the same way, descriptions of cognitive events (such as the perception of a social dilemma in a drama) may require scientific descriptions that are remote from the operation of the underlying biological parts. On the other hand, the two levels, although distant in description, are inseparable: the breaking of a single transistor can kill the display of the
biological parts and pieces. Many of the operating parts in a cell function for the metabolism, scaffolding, or reproduc- tion of the cell. Many cells may function only in supportive or nutritive roles. Our goal is to avoid a phonebook of detail in favor of teasing out the underlying principles. The details are essential for understanding how the parts of the brain work; however, simply understanding what the parts are, even in detail, is not equivalent to understanding how the parts and their interactions embody and process information. This latter problem involves understanding what things do, instead of simply what they are (FIGURE 1.5).
Which Parts Matter? In modern computers, great care has been taken to separate information content from its physical embodiment. In the brain, the clean hardware/software distinction is mislead- ing. The division of hardware and software in the brain is un- clear, if the distinction exists at all. For a complete view of cognitive neuroscience, we will explore topics from the level of single molecules to systems of neurons.
To understand the levels problem, consider the screen on your cell phone. Suppose that no one knew how a cell phone worked, but you were interested in finding out how the com- ponents of the screen gave rise to the funny YouTube videos that you watch on it. So you study the personalities of the
FIGURE 1.5 (a) The 20 amino acids that make up every protein in your body and (b) a leaf of Sanskrit manuscript. Knowing only the pieces is insufficient— as insufficient for understanding how biology functions by dint of protein machines as it is for reading a manuscript in an unknown language. The mission of this book is to provide a principles-based approach to understanding what things do.
(a) (b)
HYDROPHOBIC HYDROPHILIC
Nonpolar Non-charged sidechains Charged sidechains
Acidic
Basic
01-Eagleman_Chap01.indd 7 02/11/15 3:09 pm
8 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 8 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The outcome is that our possible thoughts and actions— the full reach of our cognition—need only to have served the re- productive success of our progenitors. Our possible thoughts and actions do not necessarily equip us with a mental appara- tus appropriate to have the right intuitions about the world— or even about our own brains.
Many aspects of neural function are adaptations that were advantageous, given the survival demands placed on our ancestors. Our psychology is no exception to this: it is also an adaptation. It is a construction that made our ances- tors reproductively successful and is not simply a picture of the physical world “out there.”
There is something strange and immaterial about our thoughts and perceptions. They appear from nowhere; they are vivid while present; and we have no access to where they were before they appeared. Stranger still are the feelings and sensations that attend our thoughts. Our conscious sense of ourselves provides no extra insight into these matters, but
funny video. How small parts can unwittingly accomplish functionality at an entirely different level is the surprising and magnificent canvas on which cognitive neuroscience is painted.
Your thoughts, decisions, and moods are underpinned by physical stuff. We know this because alterations to the brain change the kinds of thoughts we can think. During dream sleep, there are unbidden, bizarre thoughts. During the day we enjoy our normal, well-accepted thoughts, which people enthusiastically modulate by spiking the chemical cocktails of the brain with alcohol, narcotics, cigarettes, coffee, or physical exercise. The state of the physical material determines the state of the thoughts.
The physical material is absolutely necessary for normal thinking to tick along. If you were to injure your pinkie in an accident, you’d be saddened, but your normal thinking would be no different than if you had not injured your pinkie. By contrast, if you were to damage an equivalently sized piece of brain tissue, this might change your capacity to un- derstand music, name animals, see colors, judge risk, make decisions, read signals from your body, or understand the concept of a mirror—thereby unmasking the strange, veiled workings of the machinery beneath (FIGURE 1.6). Our hopes, dreams, aspirations, fears, comic instincts, great ideas, fe- tishes, senses of humor, and desires all emerge from this strange organ—and when the brain changes, so do we. So although it’s easy to intuit that thoughts don’t have a physical basis, that they are something like feathers on the wind, they in fact depend directly on the integrity of the enigmatic, three-pound mission control center.
What Is the Brain For? The brain is an evolved biological organ. As such, its products—our thoughts, actions, emotions, moods, fears, etc.—are shaped by evolutionary pressures. As the biologist E. O. Wilson writes,
The essence of the argument, then, is that the brain exists because it promotes the survival and multiplica- tion of the genes that direct its assembly. The human mind is a device for survival and reproduction, and reason is just one of its various techniques.
—On Human Nature, 1978
A device for survival and reproduction? The surprising char- acter of this observation derives from the fact that what we do think, don’t think, and possibly can’t think are all con- structions of a long, undirected evolutionary process. Some of these constructions may have arisen in response to survival pressure; that is, they are psychological adaptations— mechanisms that on average enhanced the reproductive suc- cess of those creatures that possessed them. Others may have simply arisen as neutral changes and come along for the ride.
FIGURE 1.6 Three patients with deficits in conscious visual perception as a result of injuries to specific regions of the brain. (a) The injured regions of the brain in each patient. (b) The corresponding diagrams show the deficits in each patient’s visual field (gray) resulting from the injury. Indicator arrows added by Oxford University Press.
(a) (b)
90º
270º
0º 180º
90º
270º
0º 180º
90º
270º
0º 180º
L R
BT
HK
WF
01-Eagleman_Chap01.indd 8 02/11/15 3:09 pm
How We Know What We Know 9
# 158305 Cust: OUP Au: Eagleman Pg. No. 9 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
How We Know What We Know On close inspection, even a single cubic millimeter of brain reveals itself to be a tangle of neural connections more daunt- ing than the alleyways of any medieval city. W hen taken as a whole, the hundred-billion-neuron labyrinth of the human brain might seem beyond our navigational powers altogether. Yet this is the task before us if we ever hope to answer the Big Questions of neuroscience: to tease out the mechanisms by which we pick a familiar face out of a crowd of hundreds of strangers, regale our friends with stories recounted at 250 words a minute, place ourselves instantly in the recollected scenes of a childhood decades past, choose between life’s guilty pleasures and virtuous chores, or smile at the beauty of the morning mist rising from some mountain lake at dawn.
At first glance, it may seem astonishing that we have made any progress whatsoever in linking the rich complexity of human mental life to the even more dizzying intricacy of the human brain. Fortunately, over the past 150 years, we have gradually assembled a whole toolbox of complementary research methods for linking brain and behavior. A lthough none of these research methods is completely foolproof in isolation, when used together, they have been tremendously useful in helping us to understand how the circuits of the human brain give rise to human abilities.
These tools run the gamut from detailed 19th-century microscopic observations of brain anatomy to far more elab- orate 21st-century techniques that involve inserting artifi- cial genes into specific types of neurons so that they can be turned on or off with colored pulses of light (as in FIGURE 1.7).
only amplifies the foreignness of our internal experience. W hen we think of our home address, the sensation of eating fine foods, or our feelings during a car wreck, we have no access to the actions of our nervous system associated with the thoughts (Eagleman, 2011). Complicating things fur- ther, thoughts are private, and distinct thoughts move about with one body and attend the operation of a particular brain. A lthough we may recognize that our intact body and brain are required to sustain thoughts, such an insight does not ex- plain them. Upon casual self-inspection, the mind seems quite strange and nonphysical, perhaps resisting modern sci- entific inquir y. Despite these immaterial qualities, all em- pirical evidence available today supports the idea that unthinking physical processes in our nervous systems gener- ate our thoughts (Becchio, Manera, Sartori, Cavallo, & Castiello, 2012; Lakoff, 2012; Narayanan, 2003).
The idea that our mental lives arise from the operation of trillions and trillions of unthinking parts is unsettling. It is unsettling because we identif y so completely w ith our thoughts, feelings, and behaviors—but we rarely bother to question their origins. Many people suppose that the rich and subtle cognitive landscape experienced by humans cannot be captured solely in the operation of biological parts. How could the movements of an athlete, the brush strokes of an artist, or the abstract ideas of a philosopher result from the operation of mindless parts? The domain of biological operations using cells and molecules seems so completely different from complex behaviors and mental events that some do not even hope to explain one in terms of the former. To some, it seems presumptuous to try.
Certain qualities of our perceptual experience rein- force the apparent separateness of the physical and mental. For example, although our experiences move around with our body, they don’t seem to ex ist in some specified loca- tion. Our experiences also have a k ind of special character, feeling to us like independent entities. The effortlessness of perception and action highlights their oddness. W hen we want to see, we simply open our eyes. We hear, smell, breathe, move, taste, env y, and love w ithout even tr y ing. These obser vations and their obvious appeal have caused some people to question deeply the physical origins of our mental life, both conscious and unconscious. Perhaps, they say, our psycholog y arises as a k ind of independent prop- erty of the material world, not describable by the operation of physical mechanisms. If true, this would place at least par t of our minds outside the f ield of v iew of modern science (Lakoff, 2012).
A lthough these ideas may appeal to many levels of our experiences, scientific approaches to the mind–brain prob- lem take the position that the mind is what the brain generates. These approaches assume that our experiences do result from the operation of mindless biological parts. This per- spective, as unsettling as it may be, has produced a number of different scientific approaches for studying the mind and the brain.
FIGURE 1.7 Brainbow technique. The “Brainbow” technique reveals neural circuitry in exquisite detail by causing each individual neuron to produce a slightly different mixture of fluorescent proteins, thus causing them to glow in different colors when illuminated.
01-Eagleman_Chap01.indd 9 02/11/15 3:09 pm
10 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 10 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
In broad terms, however, all of the tools in our toolbox for linking brain circuits to behavior can be sorted into four cat- egories: connectional methods, correlational methods, lesion methods, and stimulation methods. Let’s do a quick overview of each of these approaches.
Connectional Methods In the social life of neurons, you are who you know. Within the vast network of the brain, the role of any given neuron depends in large part on its inputs and outputs. For example, if a neuron sends direct output to the muscles surrounding the eye, it can be inferred to play a role in controlling eye movements. If that neuron’s inputs come from motion- sensing visual regions of the brain, we have a hint that this neuron could help to aim the eye at any sudden movements that might occur in our visual surroundings. In contrast, if its inputs come from auditory regions of the brain, this
FIGURE 1.8 Photomicrographs showing connection pathways in the brain of a rhesus monkey. These are identified by injecting a radioactive tracer substance into one brain region (indicated by the arrow in (D), allowing it to be transported along the input and output pathways of this region, and then examining the neurons and pathways elsewhere in the brain that become labeled by the tracer.
FIGURE 1.9 Diffusion tensor imaging (DTI). DTI uses an MRI scanner to create images sensitive to the diffusion of water molecules through the structures of the brain. These images are then reconstructed into maps of the brain’s connection fibers.
might instead suggest that the neuron plays a role in aiming our eyes to spot the sources of sounds that might occur around us. Or, if its inputs come from the movement- and balance-sensing organs of the inner ear, this might give us a hint that the neuron plays a role in keeping our eyes stable as our bodies jostle their way from point A to point B—a sort of camera-steadying mechanism for our visual system.
A wide variety of methods are available for our use in tracing the connections to and from a given neuron or a given region of the brain. Some involve injecting a tracer substance into the region. Certain kinds of tracer substances are taken up by the neurons and transported along the input or output tracts of neurons until they reach the final input or output terminals of the neurons themselves (FIGURE 1.8). A fter injecting the tracer substance, we can examine the brain either with the naked eye or under a microscope to see where the tracer has spread, thus obtaining a map of the inputs and outputs to the region we are studying.
This approach has been in use for well over a century and was indispensable to early neuroanatomists who sought to make sense of the vast number of interconnecting pathways within the nervous system (Lassen, 1974). Of course, there was one major drawback to this approach: they had to actu- ally remove the brain to see where the tracers had gone. This meant that connectivity studies had to be performed either in animal species or in anatomical specimens of human brains donated after death. Yet the brains of even the most closely related animal species have important structural dif- ferences from the human brain, and postmortem human brains are often affected by disease, decay, or the effects of aging (Buxhoeveden, Lefkowitz, Loats, & Armstrong, 1996; Johnson, Morgan, & Finch, 1986).
A more recent technique, called diffusion tensor imag- ing (FIGURE 1.9), has enabled us to map out connection
01-Eagleman_Chap01.indd 10 02/11/15 3:09 pm
How We Know What We Know 11
# 158305 Cust: OUP Au: Eagleman Pg. No. 11 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
approaches are neuroimaging techniques, which revolution- ized the study of human brain function when they emerged in the 1980s and 1990s. Positron emission tomography involves injecting small amounts of radioactively labeled chemical compounds into the body and then mapping out their distribution within the brain. A wide variet y of sub- stances can be labeled, ranging from water or glucose (a simple sugar) to specially tailored compounds that bind only to a single type of chemical receptor within the brain. Magnetic resonance imaging (MR I) lets us view the struc- ture of the brain without exposing the individual to radia- tion and has become widely used both in clinical medicine and in research (see Research Methods, FIGURE 1.11). Func- tional magnetic resonance imaging (f MR I) uses a specific type of rapidly acquired MR I scan to generate whole-brain maps of blood flow and blood oxygenation within the brain. As neurons increase or decrease their activity levels, the changes register as changes in blood flow and oxygenation. Since these changes are quite localized, they can be used to generate maps of the neural activity that accompanies vari- ous forms of human cognition or behavior.
A variety of other MR I-based neuroimaging techniques also exist, including magnetic resonance spectroscopy (which can detect subtle changes in the concentrations of certain substances in brain tissue), arterial spin labeling (another method for measuring blood flow in the brain as an indirect measure of neural activity), voxel-based morphometry (which can measure subtle differences in the shape or thick- ness of brain structures), and diffusion tensor imaging (which, as above, can map the pathways of connection tracts within the brain). We will see examples of all of these methods in use throughout the rest of this book.
Studies using correlational research methods, and neu- roimaging in particular, have proliferated over the past quar- ter century. Headline-grabbing studies of the brain regions that activate during romantic love, musical improvisation, or
pathways in living human beings noninvasively. This tech- nique uses a magnetic resonance imaging scanner to create detailed maps of the directions of water diffusion within living tissue. In brains, the connection fibers (“axons,” de- scribed in Chapter 3) between distant regions of neurons tend to bundle together into tracts, traveling in parallel from region to region like the lanes on a major highway. Water molecules, like the cars on the highway, tend to travel more easily along the tracts than across them. By following the water molecules as they diffuse through the brain, we can create maps showing the most likely routes of the tracts themselves.
Numerous other approaches are available for mapping the connections between brain regions, all of which can pro- vide us with hints as to the functions of the regions them- selves. However, for these other connectional approaches to work, we still need to know the approximate functions of the input and output regions in the first place. In addition, the hints provided by connection studies must be confirmed or refuted by more direct observations. For this reason, our toolbox needs additional approaches.
Correlational Methods Correlational research methods involve making observations of brain activity, through various means, while an individual performs some type of behavior. For example, a correlational study might use a magnetic resonance imaging scanner to map the blood flow and blood oxygenation in the brains of individuals viewing familiar versus unfamiliar faces or rating the attractiveness of different pieces of artwork. By identify- ing regions and pathways whose activity correlates with the behavior under study, correlational research methods can also offer clues about what brain regions or brain mechanisms are important for specific aspects of human cognition.
More than a dozen different methods are commonly used to measure various aspects of brain activity during be- havior. On one end of the spectrum are invasive measures, such as recording the electrical activity of neurons via microelectrodes implanted directly in the brain during brain surgery. Other similarly invasive measures include the use of tiny microdialysis probes, which are capable of sampling the concentrations of chemical neurotransmitters directly from brain tissue, or voltammetry probes, which can detect neurotransmitter concentrations via minute fluctuations in electrical potential within the probes. Less invasive ap- proaches include electroencephalography, used since the 1920s to record electrical signals on the scalp that are gener- ated by oscillating electrical activity in nearby brain regions. Magnetoencephalography, used since the 1980s, records the even fainter magnetic fields that accompany this electri- cal activity (FIGURE 1.10).
On the other end of the spectrum, indirect methods can detect the metabolic or neurochemical products of brain ac- tivity rather than the activity itself. At the forefront of these
FIGURE 1.10 Magnetoencephalography. Two noninvasive scanners are positioned above and below her head. The neuromagnetometer measures magnetic fields produced by the neural activity of the brain.
01-Eagleman_Chap01.indd 11 02/11/15 3:09 pm
12 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 12 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
RESEARCH METHODS: Magnetic Resonance Imaging The idea of using magnetic fields for medical imaging is older than you might think. In 1881, Alexander Graham Bell (the inventor of the telephone) hurriedly built a crude metal detector to try to find an as- sassin’s bullet in the dying president James Gar field. Alas, the effor t failed. Today, however, millions of people owe their health to a much more sophisticated technique known as MRI. First devised in the 1970s, it is now one of the most powerful technologies for making images of living tissue (De Haën, 2001).
Most medical images are based on the way that different kinds of tissue absorb some kind of radia- tion (X-rays, or ultrasonic vibra- tions). But these differences are subtle, and it is often hard to dis- tinguish abnormal tissue (such as cancer) from healthy tissue. How- ever, it turns out that even subtle differences in the proper ties of a tissue, such as its water or fat content, can create major differ- ences in how the protons in the tissue behave when they are placed in a strong magnetic field. To make images that reveal these differences, we must par tially magnetize the body. This is not an easy feat, so ever y MRI device con- tains a superconducting magnet power ful enough to make steel furniture or equipment fly across the room in a blur. It turns out that the protons in the hydrogen atoms in your body have a proper ty called
“spin,” as if they were tiny globes spinning around an axis with a nor th and a south pole. When we put the protons in the strong mag- netic field of the MRI device, they tend to reorient themselves, so that their axis of spin align up with the magnetic field, almost like tiny iron filings. Since they are spin- ning at radio frequencies, millions of times a second, we can use radio waves to knock their spins out of alignment with the field, and radio antennas to record the energy they give off as they come back into alignment. It turns out that the way the protons absorb and return energy is exquisitely sensitive to tiny differences in the surrounding tissue: wateriness, fattiness, diffusion rates, even the oxygen content of the blood. We can use these differences to create ver y clear images of the various
t ypes of tissue within the body, at a resolution of less than a milli- meter. L ast, power ful computers and many level s of data process- ing and statistical analysis are used to turn these maps into de- tailed t wo- and three-dimensional images for viewing.
MRI is a flexible technique. Using different sequences of magnetic pulses and radio signals, we can generate hundreds of different kinds of scans. Some show gross anatomy clearly, whereas others highlight water-swollen areas of subtle injury or disease that would be invisible to the naked eye. Some black out unwanted signal from fat or fluid to spot hidden abnormali- ties. Others can even measure pat- terns of water diffusion or tissue elasticity.
FIGURE 1.11 Neuroimaging techniques. (a) A volunteer participant undergoing an MRI scan while performing a cognitive task. (b) An anatomical MRI of the participant’s brain (black and white) with a superimposed map of areas showing significant activation on functional MRI during the performance of the task (color).
(a) (b)
01-Eagleman_Chap01.indd 12 02/11/15 3:09 pm
How We Know What We Know 13
# 158305 Cust: OUP Au: Eagleman Pg. No. 13 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
sexual orgasm have brought neuroscience into the spotlight of public awareness. Behind the scenes, an ever-growing cast of less well- publicized but groundbreaking studies have mapped out the hidden crannies of human brain function in unprecedented detail.
However, at the same time, it is important to recognize the inherent limitations of neuroimaging and other correlational approaches to human brain function. Correlation, after all, does not equal causation any more than the crowing rooster calls up the dawn. For example, a neuroimaging study may find an area whose activity correlates with the unpleasant painfulness of a hot probe applied to the hand (FIGURE 1.12). Does this mean that this is a “pain unpleasantness” brain region? W hat if the region were simply involved in suppress- ing the urge to move the painful hand away from the probe— an urge that would grow stronger with increasing pain, thus producing an apparent representation of painfulness?
A lthough careful experimental design can rule out some kinds of spurious correlation, there is always the
(a)
FIGURE 1.12 Correlation does not equal causation. (a) hand with thermal probe and (b) activation in cingulate motor area. Brain regions activated during painful stimulation of the hand may reflect pain unpleasantness or other factors such as movement suppression.
(b)
possibility that a hidden third factor, C, lurks in the back- ground, producing an apparent connection between A and B where in fact none exists. A more reliable approach is to back up the findings of our correlational studies with the results of causational studies, which involve actively alter- ing brain activity and observing the effects on behavior. (A closely related approach is the observation of the effects of brain damage as a result of trauma or disease in humans.) These methods find expression in the final two categories of approach: lesion methods and stimulation methods.
Lesion Methods One of the oldest approaches to mapping out brain–behavior relationships involves studying the effects of brain lesions: areas damaged as a result of disease or other injury. A wide variety of events can cause localized damage to a part of the brain’s circuitry. Traumatic brain injuries from blows to the head, accidents, or wounds from bullets or other weap- ons can physically destroy a small region of the brain. Stroke, meaning either bleeding or blockage of the blood supply into a region of the brain, can also destroy brain tissue within a restricted region. Surgery, performed to remove a tumor or correct other abnormalities, often involves removing a part of the brain’s structure. Infections from viruses or other microbes can selectively damage certain parts of the brain while sparing others. Degenerative diseases such as demen- tia can also affect certain areas of the brain preferentially while leaving others more or less intact. In each of these cases, the loss of or damage to specific pathways within the brain often results in specific effects on cognition and behavior, which can be studied experimentally.
One of the oldest and most famous lesion studies was re- ported by the French neurologist Paul Broca in the 1860s. His patient, nicknamed “Tan,” had gradually lost the ability to pronounce any words other than the syllable for which he was named, although his ability to comprehend language re- mained intact. At autopsy, his brain proved to have a lesion in a specific region of the left frontal lobe—a region that has been known ever since as “Broca’s area,” which is linked to the production of language (Broca, 1861) (FIGURE 1.13).
Over the nex t centur y and a half, lesion studies in ani- mals and in human patients proved usef ul in identif y ing areas of the brain w ith cr ucial roles in v ision, hearing , movement, balance, touch sensation, memor y of life events, learning of motor sk ills, comprehension of lan- g uage, perception of motion and shape, problem solv ing , judgment, and decision mak ing. Even today, lesion stud- ies remain usef ul in understanding the role of specif ic brain regions in specif ic forms of cognition and behav ior. In the 21st centur y, lesions are t y pically mapped out in detail using M R I rather than autopsy, which allows us to draw links bet ween brain lesions and their behav ioral ef- fects in liv ing human beings. Several organizations retain large registries of patients w ith brain injuries so that the
01-Eagleman_Chap01.indd 13 02/11/15 3:09 pm
14 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 14 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
structures or damage part of one structure while leaving an- other part intact. W hen, for example, a patient with such a lesion shows a deficit in, say, the recollection of personal memories, it can be difficult to say exactly which of the sev- eral partially damaged brain structures is most crucial for the lost function.
Conversely, if a patient with such a lesion does not show an obvious deficit in memory recollection, we might falsely conclude that the damaged structure is not involved in this kind of memory. In fact, however, the structure could have had an important role in this kind of memory, but the lesion may have spared just enough tissue to allow it to remain intact.
Finally, in lesion studies, the exact nature of the deficit itself must be assessed carefully or the wrong conclusions can arise. This problem is nicely encapsulated in the old joke about the scientist who trains a frog to jump on com- mand and then removes its legs and tells it to jump. W hen the frog remains immobile, he writes a paper announcing a momentous discovery: that frogs who lose their legs become deaf! A lthough the error in the joke is obvious, similar lapses in deductive reasoning can easily occur in lesion studies if the types of cognition we are studying are not yet well understood.
Stimulation Methods Another powerful approach to understanding human brain function involves actively stimulating a given brain region or neural circuit and then observing the effects on cognition and behavior. Once again, this approach has a long history, extend- ing back into the mid-19th century. As early as 1870, the German neurophysiologists Eduard Hitzig and Gustav Fritsch found that by applying electrical current to specific brain re- gions in dogs, they could elicit movements of specific body parts on the opposite side (Hitzig, 1900). Similarly, in the 1940s, the Canadian neurosurgeons Wilder Penfield and Her- bert Jasper used electrical stimulation to map brain function in patients with epilepsy, producing detailed maps of brain re- gions responsible for movement, tactile sensation, speech, smell, and other functions (Penfield & Erickson, 1941). Today, studies using intraoperative stimulation of neurosurgical pa- tients continue to reveal new findings about functions such as voluntary control of behavior, the sensation of “willed” action, and the neural circuitr y responsible for sensations from inside the body (Lehéricy et al., 2000) (FIGURE 1.14).
In the 21st century, we also have a number of techniques available for stimulating the brain noninvasively in patients or even in healthy volunteers. One such method is called tran- scranial magnetic stimulation (TMS), first developed in the 1980s and now widely used in both research and medicine. TMS uses powerful electromagnetic coils, held against the scalp, to generate focused magnetic field pulses that pass through the skull to activate neurons directly underneath the
ef fects of injuries in dif ferent brain areas can be com- pared across groups of patients, rather than just among indiv iduals.
Today, large-group lesion studies are helping us pinpoint neural pathways involved in addiction, fear, sadness, plea- sure, empathy, the knowledge of social conventions, and our ability to understand the thoughts and intentions of other people. W hen combined with the detailed anatomical maps of neuroimaging and other correlational studies, lesion stud- ies can help to provide evidence for a causal role between a given set of neural circuitry and a specific form of human cognition or behavior.
There are at least three important caveats to lesion stud- ies, however. Lesions themselves are rarely kind enough to map neatly onto just one specific brain structure or circuit, while sparing its neighbors. More commonly, lesions are large, ragged injuries that sprawl across parts of two or three
FIGURE 1.13 The brain of the patient “Tan,” showing (a) the lesion in “Broca’s area” of the left frontal lobe and (b) a modern lesion study using MRI to localize the overlap among lesions in a series of 10 patients with damage to this region of the brain (Jakuszeit et al., 2013).
(b)
(a)
01-Eagleman_Chap01.indd 14 02/11/15 3:09 pm
How We Know What We Know 15
# 158305 Cust: OUP Au: Eagleman Pg. No. 15 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Another form of noninvasive brain stimulation, devel- oped more recently, is called transcranial direct current stimulation (tDCS). This technique involves applying two electrodes to the scalp (FIGURE 1.16), each about half the size of a credit card. Once they are attached, a small device passes a weak, constant electrical current across the two electrodes so that the current passes through the scalp and the underly- ing brain regions. Neurons under the positively charged cathode tend to be inhibited by this kind of stimulation, whereas neurons under the negatively charged anode tend to become excited. A lthough the currents are weak, even 20 minutes of tDCS can produce measureable changes in a variety of functions, depending on the site and polarity of stimulation: movement, perception, attention, memory, emo- tion regulation, impulsivity, and even deception (Fregni, Boggio, Nitsche, & Pascual-Leone, 2005; Nitsche & Paulus, 2000; Nitsche et al., 2008).
As with all the other approaches, stimulation studies have their own caveats. One of the most important is that the effects of the stimulation may spread well beyond the target site into other regions that are physically adjacent or strongly connected to one another. For example, we know that apply- ing TMS to one hemisphere causes effects not only at the site of stimulation, but also at the same site in the opposite hemi- sphere via connections that cross from one side of the brain
FIGURE 1.15 Transcranial magnetic stimulation (TMS). A TMS device generates a powerful, focused magnetic field pulse that can stimulate the brain noninvasively. The device can be positioned accurately over a target region of the brain using MRI guidance.
FIGURE 1.16 Transcranial direct current stimulation. Transcranial direct current stimulators can modulate brain activity using mild electrical currents applied to the scalp.
FIGURE 1.14 A neurosurgical patient undergoing intraoperative mapping of brain function during surgery using a stimulation electrode. Observations during these types of procedures have revealed many important aspects of human brain function over the past century.
Bipolar electrode
Stimulation site
Cortex
site of stimulation (FIGURE 1.15). TMS coils can be used to gen- erate movements of the body, just as in the studies of Jasper and Penfield. However, they are also being used to study the neural circuitry responsible for a much wider range of func- tions from vision and hearing to planning, memory, emotion regulation, social cognition, and decision making. Some patients with depression respond to repetitive transcranial magnetic stimulation (rTMS), multiple sessions of TMS.
01-Eagleman_Chap01.indd 15 02/11/15 3:09 pm
16 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 16 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
our brains to inch their way forward, discovery by discovery, toward understanding their own workings.
Is the Brain equipped to Understand Itself? One of the strangest features of the enterprise we call neuro- science is that it is possible at all. A lthough we rarely think about it this way, the unspoken premise of the neuroscientist is that the three-pound lump of universe within our skulls has the right kind of architecture, mechanisms, functions, and capabilities to construct an accurate, useful model of its own architecture, mechanisms, functions, and capabilities.
But is the brain really equipped to understand itself? Certainly the odds would seem to be stacked against it. A fter all, the brain is far from a universal, all-purpose computing device designed to crack any problem that falls within its sights. On the contrary, the brain is the end result of millions of generations of careful optimization toward solving three very old, very fundamental, and very specific problems (FIGURE 1.17). The first is homeostasis: keeping the body fed, watered, and generally within a happy range of survival pa- rameters. The second is agonistic behavior: defending its own survival interests against other organisms, fending off challenges from predators and rivals, and chasing down prey that made plans other than being eaten. The third is repro- duction: making sure that it leaves behind other organisms similar to itself since brains that skip this last task tend to go out of circulation rather quickly.
With the bar for survival being raised every generation, after half a billion years, brains have ended up with exquisitely
to the other. TMS of movement-controlling areas in the brain may also spread via long descending pathways to neu- rons several feet away, down in the spinal cord. So when we observe an effect from brain stimulation, we still need to clarify whether this is a direct effect from the area being stimulated or a secondary effect from stimulation spreading to areas that can be quite distant from the area originally targeted.
A Toolbox of Complementary Methods As we have seen, we can use a wide variety of approaches to tease out the complex relationships between brain and be- havior. Each has its strengths and also its weaknesses. Con- nectional studies hint at function, but only when the roles of input and output regions are well understood. Correlational studies provide detailed observations of brain activity ac- companying behavior, but do not establish a causal connec- tion between the two. Lesion studies can show a clear connection between lost circuits and lost functions, but finding “clean” lesions of individual structures can be diffi- cult, and interpreting the deficits can be even more difficult, although emerging techniques, including genetic manipula- tions, offer hope that such clean lesions may be possible soon. Stimulation studies can provide a causal link between brain activity and function, but the stimulation can also spread to distant brain areas, muddying the picture. No ap- proach is perfect. However, the convergent evidence from each of these approaches can help us figure out the specific neural pathways that are important for a given brain func- tion and the mechanisms that operate within these pathways to make that function possible.
Thinking Critically about the Brain Connection, correlation, lesion, and stimulation approaches form the key components of the methodological toolbox that has allowed us to begin picking apart the Gordian knot of the brain. Just as important, however, are a set of critical- thinking techniques that have been carefully developed over the past several centuries with the emergence of the scien- tific method of investigation. The scientific method is a tool- box in its own right—a toolbox that has allowed us to repurpose our brains to puzzle out the structure of the solar system, the causes of deadly diseases like cholera or malaria, or the mechanisms of the brain itself. These kinds of prob- lems are far afield from the brain’s usual domain of expertise: keeping a vertebrate body alive and healthy long enough to reproduce. Let’s take a closer look at the toolbox that allows
FIGURE 1.17 The critical survival functions of the brain: (a) homeostasis, (b) agonistic behavior and (c) reproduction.
Homeostatic Agonistic Reproductive
Maintaining the body’s balance of energy, temperature, hydration, and other critical parameters for one’s own survival
Defending against hostile rivals or predators, establishing territory or dominance over rivals, and seeking out prey to sustain one’s own survival
Seeking out mates, procreating, and promoting the survival of one’s own offspring
(a) (b) (c)
01-Eagleman_Chap01.indd 16 02/11/15 3:09 pm
Thinking Critically about the Brain 17
# 158305 Cust: OUP Au: Eagleman Pg. No. 17 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
beliefs. Confirmation bias is widespread in human thinking: reinforcing our prejudices about the people around us, shor- ing up our political and philosophical convictions, feeding our paranoid fantasies, and even leading physicians to de- velop false beliefs about the effectiveness of the treatments they administer (Downar, Bhatt, & Montague, 2011). It also sabotages our efforts at logical inference. For a classic exam- ple of conf irmation bias, consider the Wason card task (developed in 1966). Look at the cards in FIGURE 1.18 and decide which card or cards you would have to turn over to test the proposition, “If a card has a vowel on one side, then it has an even number on the other side.”
Intuition leads most of us to check the “A” and the “2” card. In fact, our tendency to reach for the “2” card is an example of confirmation bias. If it carries a consonant, it is not relevant to the proposition; if it carries a vowel, this supports the proposi- tion but still does not prove it conclusively because we are still in the dark about the “5.” Only by checking the “5” for vowels do we have a chance to refute the proposition by disconfirma- tion—a point we’ll return to soon. The take-home point is that, in checking our beliefs, we all too often cherry-pick the evi- dence that supports our idea while neglecting the painstaking search for the piece of evidence that would prove us wrong.
The mapping out of human cognitive biases is a major re- search area in human psychology, with more than a hundred different varieties of bias named in the past half-century alone (Barnes, 1984; Haselton & Nettle, 2006; Mineka & Sutton, 1992). In trying to understand itself, the brain must contend not only with anchoring and confirmation, but also with the availability heuristic (where scenarios feel more likely when they are more easily recalled), the affect heuristic (where the brain substitutes the easy question “How do I feel about it?” for the harder question “W hat do I think about it?”), illusory correlation (the tendency to perceive a relationship between events that are not actually connected), belief bias (in which valid arguments with hard-to-believe conclusions are re- jected), and dozens of other kinds of distortion that arise from applying a survival-oriented brain to problems outside its
complex mechanisms for meeting these challenges more ef- fectively than the competition. The brain is an organ honed to a specific set of functions, just as the lungs are honed to managing the demands of respiration and the kidneys are honed to managing the balance of water and electrolytes. Imagine how astonished we would be to find a pair of lungs dabbling in metallurgy on the side or a pair of kidneys trying to develop the perfect recipe for chili. A fter millions of years of vertebrate evolution, it should be no less surprising to find one species of brain taking time off from its usual brainy business to work on a strange new hobby: “decipher your own inner workings.”
Even if a brain should for some reason develop a yen for understanding itself, why should there be any guarantee that it is even up to the task? If the sodium-balancing functions of the kidney are ill-suited to cooking up a tolerable chili, then are the food-spotting and threat-escaping functions of the brain any better suited to cooking up a tolerable neurosci- ence? The bad news is that human cognition is full of biases and pitfalls that make the enterprise of science far from in- stinctive. The good news is that we have managed to develop a reliable toolbox of techniques for making the best of our available cognitive strengths, routing around the worst of the pitfalls, and working our way toward a useful model of our own brain functions.
Biases and Pitfalls in human Cognition As we will see in the later chapters of the book, the human brain is prone to an extensive set of biases and pitfalls that lead us to draw erroneous conclusions from our observations of the world. Worse yet, we tend to cling to our mistaken be- liefs with a confidence that far outstrips the weight of evi- dence supporting them. We’ll see these biases in more detail in Chapter 12, on human decision making, but let’s take a quick look at a couple of common ones here.
The anchoring bias describes the human tendency to become overly influenced by a single observation, usually the first observation (the “anchor”), so that it drowns out or even distorts subsequent pieces of information to make them more consistent with the anchor. A bad first impression at a job interview, an inflated price for a house on the real estate market, or a mistaken initial diagnosis in an emer- gency room all tend to take on a life of their own, becoming hard to erase even when new information comes to light. In one classic illustration, students were asked to write down the last two digits of their Social Security number before being asked to bid on a series of consumer products of uncer- tain value. Those with higher numbers bid substantially higher prices for the items, under the influence of the irrele- vant anchor number (Ariely, Loewenstein, & Prelec, 2006).
Confirmation bias is the tendency to seek out or em- phasize information that fits with our existing beliefs, while ignoring or discounting information that contradicts our
FIGURE 1.18 The Wason card task. Developed in 1966, this task is a classic example of confirmation bias. Which cards would you need to turn over to test this proposition?
“If a card has a vowel on one side, then it has an even number on the other side.”
01-Eagleman_Chap01.indd 17 02/11/15 3:09 pm
18 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 18 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
predictions, we put them to the empirical test. As with the observation phase, the experimental phase requires careful measurement and comprehensive efforts to control for any additional factors that could influence the observations, thus interfering with the predicted effects. Devising a good ex- periment also calls for a good deal of creativity, both in terms of the methods of carrying out the experiment and in terms of their execution (FIGURE 1.19). Again, breakthroughs in sci- ence often come from spotting an elegant way of testing a prediction that had previously been difficult to translate into an actual experiment.
A key point about the scientific method is that it is itera- tive: the results of one experiment count as observations that can then be fed back into the process to generate new hypoth- eses, predictions, and experimental tests. The progression, or evolution, of scientific knowledge arises from successive iter- ations of experiments, just as evolution in nature arises from successive iterations of living organisms. Over the genera- tions, as with biological systems, scientific hypotheses can become optimized into intricate, elaborate, and highly effec- tive mechanisms for addressing a given problem.
Outside this core methodology, there are some addi- tional tricks and techniques that have been added to the toolbox to circumvent the biases of our all-too-fallible human brains. A n important one is replication and exten- sion of findings: confirming preliminary findings with ad- ditional observations by other methods, or other groups of investigators, often using more rigorous approaches or more careful controls or a larger number of overall observations. Since individual observations are more likely to be distorted by extraneous factors, seeing the same results consistently across multiple tests can increase our confidence in the find- ings. The mathematical tools of statistics also help us to achieve unbiased estimates of the true values of the phe- nomena we are observing, of the likelihood that our obser- vations could have arisen by chance, and of the strength of associations or other relationships between different phe- nomena under observation.
Finally, to circumvent the influences of emotion, expec- tancy, and other biases upon the brains of individual investi- gators, the tradition of peer review brings other brains into the mix to evaluate scientific work. Independent peer re- viewers who lack a personal investment in the work are often in a better position to evaluate the validity of any given study, look for evidence of biased thinking or observation, spot missed assumptions or hidden pitfalls in the interpretation of the data, and steer clear of false implications from any findings that emerge. Traditionally, peer review is a precon- dition for publication of the results in the larger body of sci- entific literature. This quality control measure can help to ensure (but certainly does not guarantee) that knowledge disseminated through the scientific literature is less dis- torted by the common biases and pitfalls of human thought. A lthough far from perfect, this approach has helped us to make at least some headway toward understanding the workings of our own nervous systems.
usual scope of operations. Seeing the universe clearly through such distorted lenses is no easy feat. Fortunately, we also have at our disposal a toolbox of critical-thinking techniques that can correct, or at least steer our vision around, some of the worst of the distortions. It is thanks to these techniques, as much as to our microelectrodes and MR I scans, that we have been able to see our own brains with any clarity at all.
A Toolbox of Critical-Thinking Techniques It is hard to imagine advances in neuroscience, or science of any kind, coming out of a brain whose imagination is driven by what comes first to mind, whose beliefs are driven by illu- sory correlations and patterns perceived in randomness, whose observations tend to get distorted to confirm what they think they already know, and whose conclusions are rejected when they feel wrong (even when they are actually valid). Yet over the past several centuries, a new tradition of inquiry has gradually emerged to approach the universe’s many puzzles in a more systematic way: the scientific method.
At its core, the scientific method has four main steps. The first is to begin with an observation of some kind: typically an observation that is puzzling, or novel, or especially salient in some other way. To make further headway, ideally the ob- servation must be well characterized: careful, repeated mea- surements using standardized techniques and well-defined units of measurement help to improve the richness of the observation and ensure that it holds up against the perils of anchoring, illusory correlation, perceptual errors, and other similar pitfalls.
The next step is to develop a hypothesis: a proposed ex- planation for the observation in question. A good hypothesis should be capable of being verified, or ideally falsified (i.e., proven wrong), through some sort of experimental test. Useful hypotheses also tend to be parsimonious (i.e., as simple as possible while still fitting the observation), fit with the existing knowledge base, and be generalizable to other similar classes of observations.
The third step is to generate specific and testable predic- tions from the hypothesis: if our hypothesis X is true, then we ought to be able to observe other phenomena A, B, and C. Negative predictions are especially helpful: if our hypothesis X is true, we should not see phenomena D, E, or F. Finding D, E, or F would disconfirm our hypothesis and force us to re- consider. Generating useful predictions from a hypothesis requires not only the techniques of logical deduction, but also a healthy dose of creativity. Some of the greatest break- throughs in science involve not merely coming up with a clever hypothesis, but also spotting some key implication of the hypothesis that had eluded everyone else in the field. This implication can then be used as the basis for an experiment.
Experimental testing of the predictions is the last of the four key elements. Here, rather than relying solely on our in- tuitions about the plausibility of the hypothesis or on our even more distorted intuitions about the plausibility of the
01-Eagleman_Chap01.indd 18 02/11/15 3:09 pm
The Big Questions in Cognitive Neuroscience 19
# 158305 Cust: OUP Au: Eagleman Pg. No. 19 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
century? The list is lengthy, but over the course of this book we’ll touch on at least a dozen of them. In this section, we’ll take a brief look at what some of these questions are and why the answers are important. As you read through each chapter of the book, try to keep the corresponding questions in mind.
Why have a Brain at All? (Chapter 2) As human beings, we are used to working with the assump- tion that we have a specific body system, the nervous system, that looks after our perception, thought, and behavior. We are also used to the idea that this system is not evenly dis- tributed throughout our body like an immune system or a network of blood vessels, but instead has a “top-heav y” con- centration into a single organ, the brain, that lies at one end of the body.
But why should we do it this way? A fter all, there are entire categories of multicellular organisms that seem to get along fine without a nervous system of any kind: plants, fungi, and many members of the animal kingdom as well (FIGURE 1.20). W hat is it about our particular branch of the evolutionary tree that makes having a nervous system worth- while? Furthermore, why concentrate the nervous system into a brain, where its essential functions become vulnerable to strokes, tumors, blows to the head, or other injuries? Other animals like starfish or jellyfish don’t have these wor- ries: their nervous systems are distributed throughout their bodies, like a neural net.
In fact, there are some specific advantages to having a nervous system and to concentrating this nervous system at
The Big Questions in Cognitive Neuroscience You’ll often hear it said that “we still know almost nothing about how the brain works.” This isn’t strictly true. The fact is that over the past century and a half, thanks to a combina- tion of rigorous scientific study and a versatile toolbox of technical methods, we have learned a great deal about how the brain works. The sum of our current knowledge in neuro- science is already so large that no single human being can any longer claim to keep up with it all. Every year, some 40,000 new studies (more than 100 papers a day) are pub- lished in more than 200 different neuroscience journals. In- dividual neuroscientists can spend their entire careers focusing on the functions of a single neurotransmitter at a single receptor within a single neuron type in a single spe- cies. One of the major challenges in neuroscience today is simply finding efficient systems for indexing and organizing all the knowledge that we have already managed to acquire: intricately detailed maps of connections among brain re- gions, patterns of gene expression at different times during development, chemical signaling pathways within and be- tween neurons, or f MR I images of human brains at work.
That said, for all that we have learned, there remains an even vaster body of knowledge still to be acquired and a lengthy list of questions that remain to be answered. So what are the Big Questions in the cognitive neuroscience of the early 21st
FIGURE 1.19 The scientific method helps us to work around the in-built perceptual and cognitive biases of the brain.
Observations lead to research questions.
Findings are reviewed by peers and replicated or refuted by further experiments.
Experiments are designed to test the predictions of the hypothesis against alternatives.
Competing hypotheses lead to competing predictions.
1. Observation: Careful, repeated
measurements using standardized techniques.
3. Generation of speci�c and testable predictions:
if our hypothesis X is true, then we ought to
be able to observe other phenomena A, B, and C.
2. Development of a hypothesis:
a proposed explanation for
the observations in question; a good hypothesis should
be falsifiable.
4. Experimental testing of predictions: requires careful measurement
and comprehensive efforts to control for
any additional factors that could
influence the observations.
Overcoming bias:
Standardization of measures;
Disconfirmation of hypothesis; Peer review of
findings; Replication of results; Progress
by iterations
01-Eagleman_Chap01.indd 19 02/11/15 3:09 pm
20 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 20 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
It is likely that mental information is stored not in single cells but in populations of cells and patterns of their activity. However, it is currently not clear how to know which neu- rons belong to a particular group; worse still, current tech- nologies (like sticking fine electrodes directly into the brain) are not well suited to measuring the activity of several thou- sand neurons at once. Nor is it simple to monitor the connec- tions of even one neuron: a typical neuron in the cortex receives input from some 10,000 other neurons (Neff, 1987).
A lthough traveling bursts of voltage can carry signals across the brain quickly, those electrical spikes may not be the only—or even the main—way that information is car- ried in nervous systems. Forward-looking studies are exam- ining other possible information couriers: glial cells (the other main cell type in the brain, with functions that are poorly understood at present), other kinds of signaling mechanisms between cells (such as newly discovered gases and peptides), and the biochemical cascades that take place inside cells.
how Does the Brain Balance Stability against Change? (Chapter 4) The brain is commonly assumed to be like a fixed map, with different regions dedicated to specific tasks. But that as- sumption misses one of the most fundamental principles and powerful tools of the brain: plasticity, or the ability to change and retain the change. The brain is a dynamic system, constantly modifying its own circuitry to match the de- mands of the environment and the goals of the animal. W hereas your computer is built with hard wiring that re- mains fixed from the assembly line onward, the brain dy- namically reconfigures, ever so slightly, with each new experience. It reorganizes itself from the level of molecules
one end of the body. Knowing what these advantages are can give us some clues as to which of life’s problems our brains help us to solve. Knowing the brain’s to-do list is an essential step in understanding the functions it performs, as well as the mechanisms by which it performs those functions. We’ll look at the question of why we have a brain in Chapter 2.
how Is Information Coded in Neural Activity? (Chapter 3) Neurons, the specialized cells of the brain, can produce brief spikes of voltage in their outer membranes. These electrical pulses travel along specialized extensions called axons to cause the release of chemical signals elsewhere in the brain. The binary, all-or-nothing spikes appear to carry informa- tion about the world: W hat do I see? Am I hungry? W hich way should I turn? But what is the code of these millisecond bits of voltage? Spikes may mean different things at different places and times in the brain. In parts of the central nervous system (the brain and spinal cord), the rate of spiking often correlates with clearly definable external features, like the presence of a color or a face (Herrmann, 2001; Singer & Gray, 1995). In the peripheral nervous system, more spikes indicate more heat, a louder sound, or a stronger muscle con- traction (Letcher & Goldring, 1968; Woolf & Salter, 2000).
As we delve deeper into the brain, however, we find pop- ulations of neurons involved in more complex phenomena, like reminiscence, value judgments, simulation of possible futures, the desire for a mate, and so on—and here the sig- nals become difficult to decrypt (Elliott, Agnew, & Deakin, 2008; Fellows & Farah, 2007). The challenge is something like popping the cover off a computer, measuring a few tran- sistors chattering between high and low voltage, and trying to guess the content of the Web page being surfed.
FIGURE 1.20 An evolutionary tree of living organisms. Bilaterians, symmetrical multicellular animals with a head and a tail end, are only a fraction of the whole.
H u
m an
s B
ir d
s
Fi sh
In se
ct s
W or
m s
M ol
lu sc
s
Fu ng
i
Sp on
ge s
Fe rn
s
M os
se s
C on
ife rs
M on
oc ot
s A
m oe
bo zo
a
Ba ct
er ia
C ili
at es
A lg
ae
Bilateria
01-Eagleman_Chap01.indd 20 02/11/15 3:09 pm
The Big Questions in Cognitive Neuroscience 21
# 158305 Cust: OUP Au: Eagleman Pg. No. 21 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Why Does Vision have So Little to Do with the eyes? (Chapter 5) Vision seems effortless. You open your eyes and voilà: there’s the world, in all its beautiful shapes and colors. But in fact, vision seems effortless only because of the massive neural machiner y that underpins it—about one-third of the human brain. W hy all that neural machiner y? W hat is it all doing?
Cells in the back of the eye that sense light—k nown as photoreceptors—convert light into electrical signals. From there, the signals move through the visual system to the visual cortex. As the signals progress through the anatomy of the visual system, the degree of processing becomes more sophisticated. At low levels, neurons are responsive to spots of light in specific locations; at higher levels, neu- rons respond to, say, faces or houses in any part of the visual scene. The processing has gone from detail oriented to big picture.
But this picture of a hierarchy of processing—with sig- nals moving from the eyes to some higher level in the cortex—is not the full picture. In fact, it’s not even most of it. The story of vision is really about internally generated activ- ity in the cortex. That internal activity is sufficient for visual experience; it is merely modulated by incoming information through the eyes. This is why you can have full, rich visual experience in your dreams, even while your eyes are closed. So the visual system is not like a camera lens that simply reads in a photographic representation of the world. Instead, it is a construction of the brain that is largely built up from expectations.
how Does the Brain Stitch Together a Picture of the World from Different Senses? (Chapter 6) A ll senses start out with the same job: converting informa- tion sources from the world (say, a beam of light, a sound, a smell, or a touch) into a single, common currency that the brain can use: action potentials. Thus, the mission of the peripheral visual system is to convert (or transduce) light into electrical signals. Mechanisms in the ear convert vibrations in the density of the air into electrical signals. Receptors on the sk in and in the body convert pressure, temperature, and nox ious chemicals into electrical signals. The nose converts drifting odor molecules and the gusta- tor y system converts tastants. Each sense has separate, specialized regions of cortex devoted to processing its inputs: primar y auditor y cortex, primar y visual cortex, and so on.
in the synapses to the level of the gross anatomy visible to the naked eye (FIGURE 1.21). Without this ongoing change, there could be no learning and memory.
The principles of plasticity allow tremendous flexibility: a brain can find itself inside any body plan (four legs, two legs, wings, and so forth), and it will figure out how to config- ure itself to optimally control the muscles and limbs. The brain can find itself in any ecosystem (e.g., jungle, swamp, mountains), and it will learn how to operate in it. A brain can find itself in any country, and it will absorb the local lan- guage and culture.
An open question about change in the brain is known as the stability–plasticity dilemma: how can the brain con- stantly take in new information without interfering with what it has already stored? In Chapter 4 we will see some po- tential solutions to this dilemma. For example, some neu- rotransmitters (chemical messengers in the brain) make plasticity more or less effective so that only the most relevant information is imprinted. Moreover, there are different time scales of plasticity: some brain mechanisms change quickly, whereas others change more slowly. In this way, only the most consistent information works its way deeply into the forest of the brain.
FIGURE 1.21 Spines. Structures known as spines, which can be seen here along the dendrites of a neuron from the hippocampus, are important sites of plasticity in neural structure and function.
01-Eagleman_Chap01.indd 21 02/11/15 3:09 pm
22 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 22 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The hopes of early neuroscience—that brain areas would be understood as successive modules in a causal chain—are thwarted by the vast, loopy linkages between brain areas. Through a modern lens, brain areas are dependent for their normal functioning on the interwoven connections they ex- change. In the absence of understanding the dynamical loops, drawing conclusions about the function of a brain area from the activity of a single cell is likely to be as doomed as study- ing the global economy based on one person’s credit card report. This is why biology can be more difficult than physics. In physics, isolating a part of the system allows you to under- stand it directly; this is unlikely to hold true with brain areas.
how Does the Brain Control Our Actions? (Chapter 7) Next time you reach to pick up a coffee mug, watch your hand carefully. Were you even looking directly at the cup, or did you automatically reach for its approximate location? Are your fingers cupped to grasp the body of the mug or hooked to grasp the handle? Turn the cup 180 degrees and try reaching again: your hand will switch to the opposite grasp, without you having to think about it. Fill the cup with scalding hot liquid: your hand will now automatically switch to hold the cup by the handle, even if this entails an awkward reverse grasp. A ll of this takes place automatically, even with your thoughts directed elsewhere.
Now try something harder: try to catch the slowly grow- ing urge in your mind just before you decide to reach for the cup. W hy did you reach for it at that particular instant? W hy not earlier or later? Was there any external signal (such as a friend telling you to hurry up), or was the cue an internal one?
Now look at the larger set of plans into which these simple actions are embedded. W hat led you to sip on this particular drink in the first place? W hy did you choose this particular beverage over the available alternatives? Did you opt for this drink out of personal preference, lack of alternatives, polite- ness, or longer-term considerations of healthfulness?
Behind even the simplest of actions lies an immense com- plexity of computations, from choosing how to wrap your fin- gers around a coffee cup all the way up to deciding on the healthiest (or the tastiest) of available hot drinks to put in the cup itself (FIGURE 1.23). Within the brain, a hierarchy of control regions works to resolve our basic drives into specific desires, goals, strategies, actions, and movements intended to satisfy those drives. In Chapter 7, we will look at this hierarchy in detail and examine the mechanisms by which the brain con- trols our actions from moment to moment throughout the day.
What Is Consciousness? (Chapter 8) Think back to your first kiss. The experience of it may pop into your head instantly. W here was that memory before you
But there is a surprising property of multisensory experi- ence: despite the extensive division of labor in the cortex (vision, hearing, balance, vibration, temperature, pain, pro- prioception, and so on), none of that division is apparent per- ceptually. Instead, we enjoy a unified picture of the world “out there” instead of having a separate visual world, an auditory world, a tactile world, and so on. W hen you watch a blue bird flying to the left and squawking, the blueness and the motion and the sound do not bleed off one another, but instead appear perfectly bound together. How does the brain produce a uni- fied picture of the seen world given its specialized processing streams? This is known as the binding problem.
Although the binding problem remains unsolved, one clue is that the tight coordination of different sensory systems is un- derpinned by a rich fabric of connections—and these force dif- ferent regions to “come to agreement” in their firing patterns (Treisman, 1996). As a result of this design, one’s final percep- tion does not rely solely on the information coming in through the various channels. It depends partially on the input, but not entirely. Instead, just as we saw with vision, a good deal of what the brain believes it is experiencing depends on its expecta- tions. Sensory information does not simply pour into the body’s sensors and flow into a picture of the world (FIGURE 1.22).
FIGURE 1.22 La Familia del General by Octavio Ocampo. The same sensory input can generate two different percepts (objects that are perceived), depending on expectations. Do you see an old man clutching his coat lapel or a rustic couple standing under an archway near a dog?
01-Eagleman_Chap01.indd 22 02/11/15 3:09 pm
The Big Questions in Cognitive Neuroscience 23
# 158305 Cust: OUP Au: Eagleman Pg. No. 23 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
One of the traditional challenges to consciousness re- search is studying it experimentally. It is probable that at any moment some active neuronal processes correlate with consciousness, whereas others do not. The first challenge is to determine the difference between them. Some clever ex- periments are mak ing at least a little headway. In one of these, subjects see an image of a house in one eye and, si- multaneously, an image of a cow in the other. Instead of perceiving a house–cow mi xture, people perceive only one of them. Then, after some random amount of time, they will believe they’re seeing the other, and they will continue to switch slowly back and forth. Yet nothing about the visual stimulus changes; only the conscious experience changes (Kovács, Papathomas, Yang, & Fehér, 1996). This test allows investigators to probe which properties of neu- ronal activity correlate with the changes in subjective experience.
The mechanisms underlying consciousness could reside at any of a variety of physical levels: molecular, cellular, cir- cuit, pathway, or some organizational level not yet de- scribed. The mechanisms might also be a product of interactions between these levels. One compelling but still speculative notion is that the massive feedback circuitry of the brain is essential to the production of consciousness (Brown, 1970).
In the near term, scientists are working to identify the areas of the brain that correlate with consciousness. Then comes the next step: understanding why they correlate. This is the so-called hard problem of neuroscience, and it lies at the outer limit of what material explanations will say about the experience of being human.
how Are Memories Stored and Retrieved? (Chapter 9) W hen you learn a new fact, like someone’s name, there are physical changes in the structure of your brain. But we don’t yet comprehend exactly what those changes are, how they are orchestrated across vast seas of synapses and neurons, how they embody knowledge, or how they are read out de- cades later for retrieval.
One complication is that there are many k inds of memories. The brain seems to disting uish short-term memor y (remembering a phone number just long enough to dial it) from long-term memor y (what you did on your last birthday). Within long-term memor y, declarative memories (like names and facts) are distinct from non- declarative memories (riding a bic ycle, being affected by a subliminal message), and w ithin these general catego- ries are numerous subt y pes. Different brain str uctures seem to support different k inds of learning and memor y; brain damage can lead to the loss of one t y pe w ithout dis- turbing the others.
became conscious of it? How was it stored in your brain before and after it came into consciousness? W hat is the dif- ference between those states?
An explanation of consciousness is one of the major un- solved problems of modern science. It may not turn out to be a single phenomenon; nonetheless, by way of a preliminary target, let’s think of it as the thing that flickers on when you wake up in the morning that was not there, in the exact same brain hardware, moments before.
Neuroscientists believe that consciousness emerges from the material stuff of the brain primarily because even small changes to your brain (say, caused by drugs or dis- ease) can powerfully alter your subjective experiences. The heart of the problem is that we do not yet k now how to en- gineer pieces and parts such that the resulting machine has the k ind of private subjective experience that you and I take for granted. If we give you all the Tinkertoys in the world and tell you to hook them up so that they form a con- scious machine, good luck. We don’t have a theor y yet of how to do this; we don’t even k now what the theor y would look like.
FIGURE 1.23 Even the simplest movements are embedded in a rich hierarchy of actions, behaviors, strategies, goals, and evaluations, all under the control of specific neural circuits within the brain.
01-Eagleman_Chap01.indd 23 02/11/15 3:09 pm
24 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 24 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
questions that require caref ul consideration (Sara, 2000). We’ ll look at these issues in Chapter 9.
Why Do Brains Sleep and Dream? (Chapter 10) One of the most astonishing aspects of our lives is that we spend a third of our time in the strange world of sleep. Newborn babies spend about twice that. It is inordinately difficult to remain awake for more than a full day–night cycle. In humans, continuous wakefulness of the ner vous system results in mental derangement; rats deprived of sleep die after approx imately two to three weeks (Rechtschaffen & Bergmann, 1995). A ll mammals sleep, reptiles and birds sleep, and voluntar y breathers like dol- phins sleep with one brain hemisphere dormant at a time (Goley, 1999). The evolutionar y trend is clear, but the func- tion of sleep is not.
The universality of sleep, although it comes at the cost of time and leaves the sleeper relatively defenseless, suggests a deep importance. There is no universally agreed-on answer, but there are at least three popular (and nonexclusive) theo- ries. The first is that sleep is restorative, saving and replenish- ing the body’s energy stores (Adam, 1980). However, the high neural activity during sleep suggests that there is more to the story. A second theory proposes that sleep allows the brain to run simulations of fighting, problem solving, and other key actions before testing them out in the real world (Barrett, 1993).
A third theory—the one that enjoys the most evidence— is that sleep plays a critical role in learning and consolidating memories and in forgetting inconsequential details(Walker & Stickgold, 2004). In other words, sleep allows the brain to store away the important stuff and take out the neural trash. The emerging hypothesis is that information replayed during sleep might determine which events we remember later (Walker & Stickgold, 2004). Sleep, in this view, is akin to an offline practice session, helping to reinforce the learning process.
how Does the human Brain Acquire Its Unique Ability for Language? (Chapter 11) In the animal kingdom, human language is unique in its pos- sible methods for expressing intention, signifying identity, and broadcasting social signals. Our small planet contains more than 6,000 human languages, each of which gives speakers the ability to combine symbols in arbitrary,
Nonetheless, similar molecular mechanisms may be at work in these memory types. A lmost all theories of memory propose that memory storage depends on synapses, the tiny connections between brain cells (FIGURE 1.24). W hen two cells are active at the same time, the connection between them strengthens; when they are not active at the same time, the connection weakens. From such synaptic changes emerges an association (Bliss & Collingridge, 1993). Experi- ence can, for example, fortify the connections among the smell of coffee, its taste, its color, and the feel of its warmth. Since the populations of neurons connected with each of these sensations are typically activated at the same time, the connections between them can cause all the sensory associa- tions of coffee to be triggered by the smell alone (W hitlock, Heynen, Shuler, & Bear, 2006).
But looking only at associations—and strengthened connections between neurons—may not be enough to ex- plain memory. The great secret of memory is that it mostly encodes the relationships between things more than the de- tails of the things themselves. W hen you memorize a melody, you encode the relationships between the notes, not the notes per se, which is why you can easily sing the song in a different key.
Memor y retrieval is even more mysterious than memor y storage. W hen I ask whether you k now A lex R itchie, the answer is immediately obv ious to you w ithout your needing to consider ever y person you have ever met or heard about. There is no good theor y to ex plain how memor y retrieval can happen so quick ly. Moreover, the act of retrieval can destabilize the memor y. W hen you recall a past event, the memor y becomes temporarily sus- ceptible to erasure. Some intrig uing recent ex periments show it is possible to chemically block memories from reforming during that w indow, suggesting new ethical
FIGURE 1.24 A synapse, or point of communication between two neurons, illustrated in blue and yellow in this image. The three- dimensional structure of these neurons was painstakingly reconstructed by players of an online brain-mapping game known as eyeWire from actual two- dimensional electron microscope images of brain tissue.
01-Eagleman_Chap01.indd 24 02/11/15 3:09 pm
The Big Questions in Cognitive Neuroscience 25
# 158305 Cust: OUP Au: Eagleman Pg. No. 25 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
how Do We Make Decisions? (Chapter 12) We human beings are, in the words of one author, “predict- ably irrational” in our decision making (Ariely, 2009). Offer us a choice between $20 in a week or $30 a week after that and we tend to choose the latter. Yet, offer us a choice be- tween $20 right now and $30 in a week and we often choose the former. We pay over the odds for the chance to win large sums of money in the lottery and also pay over the odds to insure ourselves against the chance of losing large sums of money to fire, flood, accident, or theft. We spend vast sums to protect ourselves against the remote risks of a terrorist attack, yet feel blasé about the much greater risks of excessive drinking, smoking, driving without a seatbelt, or texting while driving.
W hat drives the irrational decision making of our spe- cies? To answer this question, we must understand the basic mechanisms of decision mak ing in the human brain. A lthough it is still a new f ield, the neuroscience of human decision making has already shed some light on what these mechanisms might be, as well as why they sometimes lead us to make irrational choices. In Chapter 12, we will look at some of the more common flavors of irrational decision making in human beings and the neural mechanisms that drive them.
What Are emotions? (Chapter 13) We often talk about brains as information-processing sys- tems, but any account of the brain that lacks an account of emotions, motivations, fears, and hopes is incomplete. Emotions involve physical responses to salient stimuli: the increased heartbeat and perspiration that accompany fear, the freezing response of a rat in the presence of a cat, or the extra muscle tension that accompanies anger. They also in- volve the subjective experiences that sometimes accom- pany these processes: the sensations of happiness, env y, sadness, and so on. Emotions seem to employ largely un- conscious machiner y—for example, brain areas involved in emotion will respond to angr y faces that are briefly pre- sented and then rapidly masked, even when the appearance is so short that subjects are unaware of having seen any face at all (Wink ielman & Berridge, 2004). Across cul- tures, the expression of basic emotions is remarkably simi- lar (FIGURE 1.26), and as Dar win obser ved, there are also common elements in the expression of emotions across all mammals. There are even strong similarities in physiologi- cal responses among humans, reptiles, and birds when showing fear, anger, or parental love (Ek man & Friesen, 1971; Gosling & John, 1999).
Modern views propose that emotions are brain states that quickly assign value to outcomes and provide a simple plan of action (Stein & Trabasso, 1992). Thus, emotion can
essentially infinite complexity (FIGURE 1.25). Language—our capacity to translate our inner thoughts into packets of communication—is such a fundamental part of our exis- tence that we typically only appreciate the massive underly- ing language systems when they stop functioning correctly (Mira & Paredes, 2005).
Centuries of brain damage studies—supplemented more recently by neuroimaging studies—have revealed that a network of brain areas is required to produce the various aspects of language. The complexity of the networks should come as no surprise when we consider all the aspects that comprise language. Take, for example, speech production, comprehension, repetition, semantics, syntax, feedback, and the ability to learn multiple languages that reference the same things.
Given the multifaceted, widespread brain networks, par- ticular regions of brain damage lead to particular deficits: some with the expression of language, some with compre- hension, some with recall, and so on. Models exist to explain how these brain areas interact (as we will see in Chapter 11), but these can offer only a general outline and not the particu- lars of any individual patient’s response to brain damage. The detailed story of the emergence of language from the brain remains to be solved.
And there is a related mystery. The two hemispheres of the brain are heavily interconnected, and they appear at first to be mirror images of one another. But surgeries that separate the two hemispheres have revealed that the two sides have impor- tant differences (Levy & Trevarthen, 1976, 1977). For exam- ple, the left hemisphere is usually dominant for language, whereas the right is dominant for music and spatial abilities. Such division of function between the hemispheres is called lateralization. W hy the brain is made up of two similar but nonidentical halves is still not clear. Nonetheless, it gives an understanding of why brain damage to identical areas—on the right or left sides—can cause such different results.
FIGURE 1.25 An evolutionary “tree of life” for 87 Indo-European languages.
A na
to lia
n To
ch ar
ia n
A rm
an ia
n G
re ek
A lb
an ia
n
In do
-I ri
an ia
n
Ba lto
-S la
vi c
G er
m an
ic
R om
an ce
C el
tic
01-Eagleman_Chap01.indd 25 02/11/15 3:09 pm
26 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 26 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 1.26 Facial expressions of happiness (a), sadness (b), anger (c), fear (d), surprise (e), and disgust (f) are recognized universally across human cultures.
be viewed as a type of computation, a rapid, automatic sum- mary that initiates appropriate actions. W hen a bear is gal- loping toward you, the rising fear directs your brain to do the right things (determining an escape route) instead of all the other things it could be doing (rounding out your grocery list). Emotions also have a priority-setting function that
colors the landscape of our perceptions and motivations, de- termining what is most important at any given time and what can be safely relegated to the background. W hen it comes to perception, you can spot an object more quickly if it is, say, a spider rather than a roll of tape. In the realm of memory, emotional events are laid down differently by a parallel
(f)
(a)
(c)
(e)
(d)
(b)
01-Eagleman_Chap01.indd 26 02/11/15 3:10 pm
The Big Questions in Cognitive Neuroscience 27
# 158305 Cust: OUP Au: Eagleman Pg. No. 27 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
reward mechanisms go awr y in pathologies of motivation, like addiction? Finally, is there any thing we can do to help make sure that we stick to the commitments we have made to ourselves? These are the questions we will consider in Chapter 14.
how Do I Know What You’re Thinking? (Chapter 15) In the children’s game of Rock, Paper, Scissors, the trick is to guess your opponent’s move ahead of time and choose the move that will beat it (FIGURE 1.27). The winner will reliably choose Paper just when the opponent chooses Rock, or Rock at just the right time to beat Scissors, or a surprise Scissors when your opponent confidently chooses Paper. Of course, your opponent is trying to do the same to you, so the battle of wits quickly escalates into an endless series of outguess- ing the outguesser: “he knows that I know that he knows I think he’ll choose Scissors, so he actually will choose Scissors.”
Being able to guess accurately at the thoughts and moti- vations of others is an immensely valuable faculty. It allows us not only to outfox our opponents, but also to work coop- eratively with our friends. Anticipating the thoughts and feelings of others is an essential skill for understanding jokes, avoiding faux pas, making thoughtful gestures, choosing thoughtful gifts, and navigating the complex conventions of human social behavior.
The ability to build models of the thoughts and beliefs of others, sometimes called “theory of mind,” is uniquely well developed in human beings. It almost certainly explains why we have been able to organize ourselves into the complex
memory system involving a brain area called the amygdala. In the realm of motivation, our emotional state helps to decide, literally, between apples and oranges: a choice that depends on our internal state at the time and thus cannot be resolved by intellectual means alone. Human beings also have elaborate mechanisms for regulating their emotions based on context and current behavior: pathways by which we reassure ourselves of the harmlessness of a rubber snake or understand the threat value of an apparently innocuous piece of paper that on closer inspection turns out to be a court summons.
As recently as 20 years ago, the study of emotions was relegated to the fringes of cognitive neuroscience and often considered lacking in scientific rigor. With the development of new experimental methods, however, so-called “affective neuroscience” has become one of the most active areas of re- search in the field. One goal of affective neuroscience is to understand the many disorders of emotion, depression being the most common and costly. Impulsive aggression and vio- lence are also thought to be consequences of faulty emotion regulation. In Chapter 13, we will look at what has been learned so far about the neural mechanisms of emotion and how these mechanisms can sometimes go awry in mental disorders, with catastrophic results.
how Do We Set Our Priorities? (Chapter 14) One of the curious ironies of human existence is the divide between intelligence and priority setting—the so-called “common sense” that is neither common nor, strictly speak- ing, a sense. Ever y one of us knows people who are recog- nized for their cleverness and zip through intellectual puzzles with lightning speed, yet are hopeless at prioritiz- ing activities in their lives. We wonder how someone so in- telligent could become so bogged down in irrelevant detail, bur y themselves in activities of minor importance while neglecting urgent issues, or procrastinate to the point of failure.
Nor, if we are honest, do we see these qualities only in other people. Every one of us has made a New Year’s resolu- tion only to break it later, found ourselves an excuse to put off our homework until the next day, or delayed working on an assignment even when we know it will soon be overdue. Our nearsighted view of life’s rewards all too often leaves us with a skewed set of priorities.
Yet how is this possible? For what reason can we k now, intellectually, the right thing to do and yet still fail to do it? W hy does our alarm clock never fail to activate at the ap- pointed hour, while the vastly greater processing power of our brain all too often finds an excuse to hit the snooze button? W hat can neuroscience tell us about the mecha- nisms of reward, motivation, and judgment? How do
FIGURE 1.27 Competitors in the Rock, Paper, Scissors Championships must try to outguess one another’s moves at levels consistently above chance. Although we might expect the outcomes to be random, in fact certain individuals display consistently higher-than-average abilities to guess what their opponents are planning to do.
01-Eagleman_Chap01.indd 27 02/11/15 3:10 pm
28 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 28 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
progressive loss of faculties in A lzheimer’s dementia or the stubborn persistence of joylessness and despair in major depression, we now have at least some idea of what brain circuitr y is involved and how this circuitr y might have gone awr y.
The hope, of course, is that this new knowledge will lead to new treatments for some of the world’s most devastating illnesses (FIGURE 1.28). A lthough much of this hope has yet to be realized, we do have at least a few hints as to what the next generation of treatments will look like for disorders of the brain. In Chapter 16, we will look at a sampler of illnesses across the spectrum from those traditionally considered “ brain diseases” to those traditionally considered “mind dis- eases.” We will look at what seems to have gone awry in each case, as well as what treatments might someday be developed to reverse the underlying causes of illness.
The Payoffs of Cognitive Neuroscience Neuroscience in general, and cognitive neuroscience in par- ticular, has become one of the most active areas in all of 21st-century science. Worldwide, there are now dozens of health organizations, hundreds of academic institutions, and thousands of private enterprises spending billions of dollars annually to further our understanding of the brain. The scale of these investments gives us at least a rough idea of the size of the anticipated payoffs that are expected to arise from research in cognitive neuroscience over the coming century. In this section, we’ll look more closely at what kinds of payoffs we might hope to see from advances in neurosci- ence over our own lifetimes.
social structures of modern civilization—structures that, among other things, enable us to take time out from foraging for food and shelter so that we can puzzle away at the work- ings of our own brains! The neuroscience of social cognition has quickly become one of the most active fields in the modern study of the brain. In Chapter 15, we will look at what we have learned so far about the social faculties of the human brain, which play such a pivotal role in enabling com- plex forms of human behavior.
What Causes Disorders of the Mind and the Brain? (Chapter 16) Neurological and psychiatric disorders are among the most prevalent and the most disabling of all illnesses, in the devel- oping as well as in the industrialized world. Major depres- sion, social anxiety, Alzheimer’s dementia, stroke, traumatic brain injury, epilepsy, schizophrenia, bipolar disorder, multi- ple sclerosis, and other brain diseases are not only exceed- ingly common, but also carry a collective burden of disease and disability that eclipses virtually all other forms of human illness—including perennial killers like malaria, malnutri- tion, HIV, heart disease, cancer, and death by accident or violence. According to estimates by the World Health Orga- nization, neuropsychiatric diseases will account for fully 14.7 percent of the global burden of disease by 2020 (Murray & Lopez, 1997).
The past centur y and a half, and in particular the past 50 years, have seen tremendous advances in understanding the neural basis of brain disorders. Traditionally, medicine has divided these disorders into the neurological—diseases of the ner vous system—and the psychiatric—diseases of the mind. As we have learned more about these illnesses, the distinction between the two has gradually become blurred. W hether we are look ing at the memor y lapses and
FIGURE 1.28 Neuroimaging studies are revealing the neural underpinnings of a variety of psychiatric conditions. Shown here are regions of the brain that undergo a subtle shrinkage of volume in (a) depression, (b) bipolar disorder and schizophrenia, and (c) posttraumatic stress disorder.
(a) (b) (c)
01-Eagleman_Chap01.indd 28 02/11/15 3:10 pm
The Payoffs of Cognitive Neuroscience 29
# 158305 Cust: OUP Au: Eagleman Pg. No. 29 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
mask ing sy mptoms, neuron-regrow ing treatments could replace damaged circuitr y and restore lost f unctions. A s it turns out, the brains of many species are capable of re- grow ing damaged areas, particularly if the damage occurs early in life. Even in human beings, new neurons are con- stantly produced in the brain throughout adulthood, albeit not in numbers large enough to completely heal large re- gions of injur y ( Eriksson et al., 1998; Pérez-Cañellas & García-Verdugo, 1996). If the brain’s self-repair mecha- nisms could be stimulated in adults, whole categories of currently incurable neurological illness might become temporar y inconveniences.
Brain stimulation also offers a potentially powerful new avenue for treating neurological and psychiatric disor- ders. Deep brain stimulators have now been used for several decades to treat movement disorders such as Park inson’s disease (FIGURE 1.29). More recently, these same devices have been used successfully to treat a wider variety of con- ditions ranging from depression and obsessive– compulsive disorder to the memor y dysfunction of A lzheimer’s disease (Greenberg et al., 2006; La xton et al., 2010; Mayberg et al., 2005). Less invasive forms of brain stimulation such as rTMS and tDCS (which we saw earlier in the chapter) are also emerging as treatments for a variety of neurological and psychiatric conditions. To use these treatments effec- tively, we must be able to target the right area of the brain for each condition. As we’ ll see throughout this book, on- going research into the mechanisms of depression and other brain disorders has been helpf ul in identif y ing
healing the Disordered Brain Compared to even a centur y ago, modern medicine has made tremendous strides in its ability to heal common forms of illness. Effective antibiotics and antiviral agents have driven many once-fearsome infectious diseases into the background. Advances in emergency and intensive care have vastly increased the rate of sur vival from accidents and injuries—even on the battlefield. Safe and effective an- esthesia has enabled many once-risky forms of surger y to become routine. In the developed world, death in child- birth has receded from a commonplace tragedy to a rare anomaly.
Yet for many brain diseases, effective treatments remain elusive. For elderly patients w ith A lzheimer’s de- mentia, a few medications can delay the progression of sy mptoms, but none can alter the course of the disease. For many other forms of dementia, there are still no effec- tive treatments at all. Patients who suffer strokes can sometimes, w ith great effort, shift their lost faculties to sur v iv ing brain regions. However, the damaged regions themselves are, for the moment, beyond physical repair. The same is true for spinal cord injuries, which leave their v ictims’ bodies partially or totally paralyzed. Schizophre- nia is a lifelong condition that appears in early adulthood and leaves patients w ith hallucinations, delusions, and bi- zarre patterns of behav ior severe enough to preclude normal life in the majorit y of cases. A lthough there are medications that can relieve some sy mptoms, for the moment there is no cure for the illness itself. Depression is one of the most common forms of psychiatric illness, af- fecting as many as 1 in 10 people in the developed world at any given time (Gonzalez et al., 2010). A lthough medica- tions and therapy are helpf ul in many cases, up to one- third of patients do not respond to these k inds of treatment (P. S. Wang et al., 2005). Unfortunately, despite extensive research, no new classes of antidepressant medication have been brought to market in decades. Even our ex isting antidepressant medications are scarcely more effective than the ones available in the 1950s—although the side ef- fects have improved considerably.
Given this current situation, one of the major payof fs ex pected from neuroscience research is in the develop- ment of new, ef fective treatments for disorders of the brain. For example, studies of learning and memor y have shed light on the cellular and molecular mechanisms that go aw r y in A lzheimer’s disease and other forms of demen- tia. In the years to come, it may be possible to block or reverse these mechanisms to aver t the progression of illness.
The abilit y to regrow damaged neurons would open up a whole new avenue of treatment for patients w ith stroke, spinal cord injur y, traumatic brain injuries, or degenera- tive diseases of the ner vous system. R ather than simply
FIGURE 1.29 Deep brain stimulation. A skull X-ray of a patient who has undergone surgical implantation of a pacemaker-like device known as a deep brain stimulator. These devices are used to treat an increasingly wide variety of neurological and psychiatric conditions including Parkinson’s disease, obsessive–compulsive disorder, and major depression.
01-Eagleman_Chap01.indd 29 02/11/15 3:10 pm
30 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 30 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
for memory, instructional videos for skills like unplugging a drain or making a crème brûlée, and online reference works for the vast reams of factual knowledge too large to fit inside our heads. In the years to come, neuroscience research will likely lead us to more direct interfaces that allow for more seamless connections between our brain and external devices.
Some such devices are already entering common use (FIGURE 1.30). For decades, cochlear implants have provided hearing to individuals who would otherwise lack it. More recently, retinal implants have provided a modicum of vision to patients who would otherwise be blind. These chips, implanted directly into the central part of the retina, bypass the light-sensing cells of the eye to stimulate the un- derlying neural pathways directly (Dowling, 2008). A l- though they are designed to take input from a small, eyeglasses-mounted external camera, they could be easily configured to take input from other sources: a telescope, an X-ray emitter, a night- vision camera, a mass spectrometer, a navigation system, or a Web browser. Such devices might enable whole new classes of human capability: navigating a surgical probe through hard-to-reach areas of the body under visual guidance, sensing the chemical environment of the air by sight, or overlaying “augmented reality” tags on top of the visual scene.
Memory, like any other sense, may also be amenable to interfaces. A lready, studies are underway to enhance memory function in A lzheimer’s disease through stimula- tion of the relevant neural pathways. Microelectrode record- ings are revealing how memories are encoded and retrieved as patterns of neural activity in brain structures like the hip- pocampus (Bliss & Collingridge, 1993). These patterns could conceivably be used to index and retrieve externally stored sources of information, such as electronic databases. Inter- faces for output devices are also under development. In the motor areas of the brain, implanted electrode arrays are used to allow paralyzed patients to control mouse pointers or
effective targets for brain stimulation. The enterprise of mapping out abnormally active circuits in brain disorders and then targeting them for stimulation w ill likely be one of the more powerf ul applications of neuroscience in the near f uture.
enhancing human Abilities A side from restoring lost f unctions in disease, cognitive neuroscience research also offers the possibilit y of en- hancing the ex isting abilities of the healthy human brain. In its simplest and most innocuous form, this may simply involve developing a better science of how to overcome (or at least work w ith) the predictable irrationalities of human decision mak ing. A s we learn more about what induces the brain to cheat on a diet and what induces the brain to stick to a diet, we may be able to find ways of promoting the latter and discouraging the former. Encouraging “the better angels of our nature” could involve a number of dif- ferent approaches. Brain stimulation techniques such as rTMS or tDCS could be used to strengthen pathways that promote long-term over short-term think ing or to improve our abilit y to inhibit our counterproductive impulses. New classes of medication might be found to accomplish the same goal pharmacologically rather than electrically. These sorts of tools might make it easier for us to stick by our ow n commitments: a neural aid to the New Year’s resolution.
A nother form of enhancement that is currently making the transition from science fiction to science fact is the de- velopment of human brain interface devices. A lready, we have a long history of enhancing our perceptual and cogni- tive abilities with simple aids such as eyeglasses or contact lenses for vision, public-address systems or hearing aids for hearing, thesauruses and dictionaries for language, GPS systems for navigation, electronic organizers and reminders
FIGURE 1.30 Brain interface devices. Some such devices already in use include (a) cochlear implants to provide auditory input and (b) retinal implants to provide visual input. Output devices are also entering use. (c) A miniaturized sensor array of 96 microelectrodes, less than a centimeter across, can be implanted in areas of the brain that control movements of the arm and hand. (d) A woman with complete paralysis below the neck uses this implanted device to guide a drink to her lips, using a robot arm controlled by her thoughts, via the implant.
(a) (b) (c) (d)
01-Eagleman_Chap01.indd 30 02/11/15 3:10 pm
The Payoffs of Cognitive Neuroscience 31
# 158305 Cust: OUP Au: Eagleman Pg. No. 31 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
the complex patterns of brain activity seen in functional MR I scans. In effect, this amounts to us borrowing from the functional architecture of the brain to build tools that help us better understand the functional architecture of the brain. These kinds of techniques are already powerful: in one study, researchers were able to build rough reconstructions of the visual scenes that subjects were viewing simply by analyzing their patterns of f MR I brain activity using an appropriately trained neural network (Naselaris, Kay, Nishimoto, & Gallant, 2011; more on this in Chapter 5).
Despite these advances, there remain many k inds of human cognition whose artif icial counterparts are rudi- mentar y at best: understanding the semantic content of human communication, responding to questions posed in natural lang uage, understanding others’ thoughts and in- tentions, self-awareness and self-ref lection, creative prob- lem solv ing , and the generation of emotional states. It is in such areas that an improved understanding of the under- ly ing neural mechanisms w ithin the human brain might lead to the development of artif icial mechanisms that per- form analogous f unctions in the realm of machine intelligence.
Brain-Compatible Social Policies Cognitive neuroscience enjoys increasing payoff in the domain of social policy. Let’s consider three examples.
Eyewitness Testimony At least since 1967, the U.S. Supreme Court has recognized that eyewitness identification evidence is the kind of testi- mony that “ juries seem most receptive to, and not inclined to discredit.” And yet they have also recognized that it is “noto- riously unreliable” (Young, 1981). In other words, eyewit- ness testimony is almost certainly the worst “technology” allowed in courtrooms today and yet the one with the most sway on jurors (FIGURE 1.31).
As we will see in Chapter 9, memory is not like a video recorder; instead, it is a reconstruction. More than 30 years of cognitive neuroscience studies have revealed issues that contaminate accurate memory recall, including issues such as weapon focus (concentrating on a weapon at the expense of encoding details about a perpetrator), cue overload (too many things going on too quickly), the other-race effect (it is more difficult to distinguish faces of races other than faces of one’s own race), and the detrimental memory effects of gen- eral stress and trauma during the event (Chance & Goldstein, 1996; Eakin, Schreiber, & Sergent-Marshall, 2003; Kramer, Buckhout, & Eugenio, 1990). A more general problem is called the misinformation effect: misleading information presented between the encoding of an event and its subse- quent recall influences a witness’s memory. So if you’ve seen an event and then are told about it (say, by another witness), the subsequent description will irrevocably alter your own memory.
other input devices (Suner, Fellows, Vargas-Irwin, Nakata, & Donoghue, 2005). In animals, interfaces with mechanical effectors (such as robot arms) have been under study for more than a decade, and human applications are emerging (Iáñez, Azorín, Úbeda, Ferrández, & Fernández, 2010).
Mak ing f ull use of these ex tended capabilities w ill re- quire systems that are well designed to tap into the brain’s ex isting f unctional architecture. For example, why rein- vent a memor y retrieval system when the brain already has circuitr y neatly optimized for this role? Instead, the brain’s ex isting circuitr y could be used to index a much larger database of ex periences and k nowledge. Likew ise, the brain’s highly eff icient action-g uidance systems could be used to nav igate through databases of information, control ex ternal dev ices w ith speed and precision, or solve complicated v isual–spatial problems like work ing out the three- dimensional str ucture of a biological protein—a task that currently requires enormous computational re- sources. To bring these possibilities to life, we w ill need to have a detailed understanding of the computational archi- tecture of the brain at the microscopic scale of neural net works and interconnections.
Blueprints for Artificial Cognition Another payoff from cognitive neuroscience will come from borrowing the brain’s best tricks to improve the abilities of our computing devices. There are a large number of prob- lems that are exceedingly difficult for the human-designed architecture of computer hardware and software, but are readily solved by human brains. Recognizing and interpret- ing speech, balancing a moving body on two legs, recogniz- ing thousands of different classes of objects from a visual scene, distinguishing familiar from unfamiliar faces when viewed from almost any angle, finding efficient routes through a series of destinations, retrieving relevant memo- ries from a large database: for each of these problems, evolu- tionary processes have endowed the brain with algorithms far more efficient than the best we have managed to engineer in computers.
Borrowing from the brain’s architecture has already proved useful in some forms of computation, in which math- ematical tools that resemble neural networks can be “trained” to distinguish among different kinds of inputs: for example, to turn written text into digitally encoded words, to predict the words we intend to type into an electronic device, or to spot faces in a photograph so we can remove the “red eyes” induced by a camera flash. The trained networks can be used to categorize new inputs that were not part of the original training set in a non-rule-based way that involves picking out the most relevant “features” within the incoming data, much as in a real neural network (Dony & Haykin, 1995; Golden, 1996; J. Y. Wang & Zhang, 2001).
In a strange, full-circle application of this technique, “neural network ” approaches have even been used to classify
01-Eagleman_Chap01.indd 31 02/11/15 3:10 pm
32 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 32 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
As cognitive neuroscience has unmasked these sorts of problems, it has been able to construct real-world solu- tions. Consider police lineups. Many memor y problems plague the accuracy of the eyewitness identification of a person in a lineup, including police suggestibility (in which an investigator influences an eyewitness), co-witness con- tamination (in which multiple witnesses accidentally influ- ence each other’s versions of the details), and contamination of memor y by photos subsequently seen in the media (Wells et al., 1998). The illumination of such problems has led to better guidelines for police who conduct lineups— these include using investigators who are blind to the sus- pects in the case, separating witnesses as soon as possible, and not publishing photos of suspects in the news (Wells et al., 2000).
Attack Demand, Not Supply Nearly 7 of 10 jail inmates met the criteria for substance abuse or dependence in the year before their admission.
The War on Drugs is utterly unwinnable. W hy? Because we are attacking the drug supply. W hen drug enforcement agents attack drug supply in one location, it pops up else- where. The right way to deal with drug problems is to address the demand for drugs—in other words, the brain mecha- nisms that drive addiction and make it so hard to overcome (Eagleman, Correro, & Singh, 2010). As we will see in vari- ous chapters, a good deal is understood about the circuitry and pharmacology of drug addiction. Addiction may be rea- sonably viewed as a neurological problem that allows for medical remedies, just as pneumonia may be viewed as an affliction of the lungs. As we progress in our understanding of the underlying circuitry of addiction—how that circuitry leads to drives and how drugs hijack and regulate that cir- cuitry— we have the opportunity to leverage that under- standing into more effective drug policy that rests on treatment rather than punishment.
With that understanding in place, promising new tech- nologies relating to emerging knowledge and technologies
FIGURE 1.31 Human brains tend to make rapid judgments about a person’s trustworthiness, friendliness, competence, and other personal characteristics based purely on their facial features. These baseless first impressions are surprisingly influential and difficult to overcome and can affect jury deliberations and election results alike. For example, if you had to trust one of these people to look after your home for the weekend, do you have any “gut feelings” about whom you would choose or whom you would avoid?
01-Eagleman_Chap01.indd 32 02/11/15 3:11 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 33 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Key Principles 33
may provide a bridge between the failed policies of the past and novel solutions in the future. Drug addiction is rooted in the biology of the brain, and society’s best hope for breaking addiction lies in new approaches to rehabilitation, not simply in repeated incarceration.
Implications for Criminal Punishment and Rehabilitation W hen a convicted criminal stands in front of the judge’s bench today, the legal system wants to know whether he is fully blameworthy or whether he had mitigating biological problems. In other words, was it his fault or his biology’s fault?
But this is the wrong question to be asking. The choices we make are inseparably yoked to our neural circuitry, and therefore there is no meaningful way to tease the two op- tions apart. The more society learns about neuroscience, the more the seemingly simple concept of blameworthiness be- comes nuanced, and the more our legal system has to come up to speed with the realities of the brain.
None of this means that lawbreakers will be let off the hook. Instead, neuroscience can contribute to building an evidence-based legal system that will continue to take crimi- nals off the streets, but will customize sentencing, leverage new opportunities for rehabilitation, and better understand how to structure incentives. Discoveries in neuroscience suggest a new way forward for law and order—one that will lead to a more cost-effective, humane, and flexible system than the one we have today (Eagleman, 2011).
As an example, an estimated quarter of the population in American prisons is mentally ill, which means that our prison system has become our de facto mental health–care system (Birmingham, Mason, & Grubin, 1996). As we will see in Chapter 16, there are more fruitful ways than incar- ceration to help those with mental problems.
In all the above examples, decades of cognitive neurosci- ence research in the lab have the direct chance to inspire real social change.
Conclusion We experience the world so effortlessly that it is easy to overlook the fact that we are not even remotely aware of the autonomous neural mechanisms that underlie our per- ceptual capacities and behav ior. The world of unthink ing biological mechanisms, operating w ith no apparent inten- tions or goals, seems at first glance to be insufficient to ac- count for complex behav iors and perceptual capacities. Nevertheless, all empirical ev idence available today sug- gests that the mind is what the brain generates. The human brain has remarkable capabilities that are unmatched by any thing else we k now of, from subatomic particles to supernovae.
The mechanisms that allow our little three-pound frag- ments of the Big Bang to think, sense, feel, and act are being explored in greater detail and at a faster pace today than at any other time in human history. Thanks to an ever- expanding kit of neat technical tricks and a trusty toolbox of scientific methods, our brains are making remarkable prog- ress in understanding their own workings. Many big ques- tions remain, and we will be exploring some of them in great detail throughout the rest of this book. If unraveled, these knotty problems may yield solutions with big payoffs, from healthier brains to healthier societies. So, without further ado, let’s gather up our enthusiasm, our curiosity, and our critical-thinking skills and dive into one of the last, greatest frontiers of human understanding: the inner workings of the human brain.
KEY PRINCIPLES
• The human brain is one of the most complex and remarkable objects in the known universe.
• Cognitive neuroscience seeks to understand how the brain gives rise to perception, emotion, aware- ness, memory, planning, decision making, and the many other varieties of human thought.
• We cannot understand how the brain as a whole func- tions simply by understanding its component parts.
• Studies of the brain’s structural connections, brain–behavior correlations, the effects of lesions,
and the effects of brain stimulation all help to un- ravel the mechanisms behind these cognitive functions.
• The brain is not intrinsically designed to understand its own workings and is subject to built-in cognitive biases that interfere with logical deduction.
• A toolbox of techniques known as the scientific method can help us correct for the built-in cogni- tive biases that might otherwise frustrate our ef- forts to understand the workings of our own brains.
01-Eagleman_Chap01.indd 33 02/11/15 3:11 pm
34 PART 1 • ChAPTeR 1 Introduction
# 158305 Cust: OUP Au: Eagleman Pg. No. 34 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
• We have learned a great deal about how the brain works using these methods, yet many big questions remain to be answered, including how information is coded in neural activity, how the brain stitches together a picture of the world from different senses, what consciousness is, how memories are stored and retrieved, and how we make decisions.
• Advances in neuroscience may lead to new treat- ments for brain disorders, enhanced human abili- ties, blueprints for artificial cognition, and better- informed social policies.
KEY TERMS
Who Are We? cognitive neuroscience (p. 5) emergent properties (p. 5)
In Pursuit of Principles adaptations (p. 8)
How We Know What We Know genes (p. 9) tracer (p. 10) tracts (p. 10) diffusion tensor imaging (p. 10) microelectrodes (p. 11) microdialysis (p. 11) voltammetry (p. 11) electroencephalography (p. 11) magnetoencephalography
(p. 11) positron emission tomography
(p. 11) glucose (p. 11) magnetic resonance imaging
(MRI) (p. 11)
functional magnetic resonance imaging (fMRI) (p. 11)
voxel-based morphometry (p. 11)
lesions (p. 13) traumatic brain injuries (p. 13) stroke (p. 13) tumor (p. 13) transcranial magnetic
stimulation (TMS) (p. 14) repetitive transcranial
magnetic stimulation (rTMS) (p. 15)
transcranial direct current stimulation (tDCS) (p. 15)
cathode (p. 15) anode (p. 15)
Thinking Critically about the Brain homeostasis (p. 16) agonistic behavior (p. 16) reproduction (p. 16)
anchoring bias (p. 17) confirmation bias (p. 17) availability heuristic (p. 17) affect heuristic (p. 17) illusory correlation (p. 17) belief bias (p. 17) scientific method (p. 18) hypothesis (p. 18) replication (p. 18)
The Big Questions in Cognitive Neuroscience
plasticity (p. 20) binding problem (p. 22) short-term memory (p. 23) long-term memory (p. 23) declarative memories (p. 23) nondeclarative memories
(p. 23) lateralization (p. 25)
REVIEW QUESTIONS
1. How would you explain the term “cognitive neu- roscience” to another person?
2. What categories of research methods are used to study human brain function? What are some examples, advantages, and disadvantages of each?
3. What are three key functions that the brain has evolved to perform as a biological organ?
4. Describe three common forms of cognitive bias in human thinking. Can you think of an example of each type in everyday life?
5. What are the key elements of the scientific method? How might each of these elements help to overcome some of the brain’s built-in biases?
01-Eagleman_Chap01.indd 34 02/11/15 3:11 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 35 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical-Thinking Questions 35
CRITICAL-THINKING QUESTIONS
2. In what ways could advances in neuroscience provide benefits for human society? Can you think of two potential dangers posed by ad- vances in neuroscience? How might we address these dangers in a constructive way?
1. What are some of the outstanding questions in cognitive neuroscience? Can you think of two important questions about the brain other than the ones that appear in this chapter?
01-Eagleman_Chap01.indd 35 02/11/15 3:11 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 36 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Describe the basic underlying organization of all
vertebrate central nervous systems.
• Summarize the basic organization and structure of the peripheral nervous system.
• Explain the circuitry and function of spinal reflexes and central pattern generators.
• Distinguish the major components of the brainstem and their functions.
• Characterize the anatomy of the cerebellum and its role in motor function.
• Illustrate the role of the hypothalamus in homeostasis and the role of the thalamus as a relay and synchronization center, using examples.
• Identify the locations of the four lobes of the cerebral cortex, the locations of the major gyri and sulci, and their functions.
• Characterize the components of the basal ganglia and their functions.
• Distinguish the major components of the limbic system and their functions.
36
02-Eagleman_Chap02.indd 36 02/11/15 3:15 pm
37
# 158305 Cust: OUP Au: Eagleman Pg. No. 37 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The Brain and Nervous System
STARTING OUT: The Brains of Creatures Great and Small
An Overview of the Nervous System The Peripheral Nervous System The Spinal Cord
CASE STUDY: Christopher Reeve, 1952–2004
THE BIGGER PICTURE: In Search of a Cure for Spinal Cord Injury
The Brainstem
NEUROSCIENCE OF EVERYDAY LIFE: Why Do We Get the Hiccups?
The Cerebellum The Diencephalon: Hypothalamus and Thalamus
CASE STUDY: Waking the Brain
The Telencephalon: Cerebral Cortex and Basal Ganglia
RESEARCH METHODS: Cytoarchitecture of the Cortex
Uniting the Inside and Outside Worlds
CHAPTER 2
02-Eagleman_Chap02.indd 37 02/11/15 3:15 pm
38 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 38 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT: The Brains of Creatures Great and Small The smallest mammal on Earth is barely larger than a bumblebee. The hog-nosed bat, which lives in remote areas of Myanmar and Thai- land, weighs a mere 2 grams and has a brain that is only a quarter of an inch in length. Yet, this tiny brain can accomplish a remarkable range of feats. It enables the bat to hover and fly, steering flawlessly through three-dimensional space, avoiding obstacles along the way. It equips the bat to pluck insects out of the air and continue onward without breaking its flight. Moreover, it
allows the bat to find its prey in total darkness by echolocation: high- frequency sound pulses whose au- ditory reflections reveal an object’s direction, distance, and movement. At the same time, the bat’s tiny brain can still process information from our more familiar senses: vision, touch, smell, taste, balance, and position sense. In addition to all of this, the bat’s nervous system is highly involved in performing es- sential tasks of survival, including respiration, hormone regulation, temperature regulation, shelter
seeking, mate finding, and rearing of offspring. All in all, this is an im- pressive set of skills for a nervous system smaller than a raisin.
Another echolocator is one of the largest mammals on Earth: the adult male sperm whale. This 50-ton predator can be found more than a mile below the surface of the ocean, diving for up to 45 minutes at a time, cruising the depths of the ocean in search of the half-ton of fish and squid it must consume every day to survive. Its brain weighs nearly 20 pounds and is the
FIGURE 2.1 Common structure. There is a common structure to the brains of all vertebrate and invertebrate animals, from mammals like the hog- nosed bat (a) or the sperm whale (b), to ancient, jawless fishes like the hagfish (c), to insects like the bumblebee (d).
(d)(c)
(a) (b)
02-Eagleman_Chap02.indd 38 02/11/15 3:15 pm
An Overview of the Nervous System 39
# 158305 Cust: OUP Au: Eagleman Pg. No. 39 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
size of a large Thanksgiving turkey. It continues to function at depths where the water pressure exceeds 3,000 pounds per square inch.
The adult male sperm whale and the hog-nose bat live in vastly different environments, having gone down separate evolutionary paths more than 80 million years ago. However, if we were to assume that their nervous systems are dramatically different from one an- other in structure and function, we would be wrong.
In fact, the brains of all mam- mals have a common underlying structure, whether they belong to bats, whales, or humans (FIGURE 2.1). The basic architecture of the
nervous system emerged so long ago that we can find much of it even in such distant relatives as the hagfish: a strange, jawless, spine- less, deep-water scavenger capa- ble of tying its own body into a knot. Its last common ancestor with the sperm whale lived roughly half a billion years ago, yet its brain has a similar underlying structure. Even the brains of insects have underly- ing similarities to our own, espe- cially in the genes that guide their development. When the genetic signal that tells a mouse where its eye needs to develop is placed into a fruit fly embryo, the fruit fly will develop an eye in that location (a fruit fly eye, not a mouse eye!).
The underlying similarities of all nervous systems should remind us that the human brain is an old, old heirloom. Although much of this book will focus on the cognitive abilities of human beings, we must keep in mind that these abilities arise from a nervous system whose fundamental organization is an- cient and has been highly con- served over time. In this chapter, we will take a first look at the struc- ture and function of the human nervous system from one end to the other. Along the way, we’ll point out both the common features and the differences that have emerged between our relatives and us over the past 500 million years.
An Overview of the Nervous System
Why Put Your Neurons in a Brain at All? The first striking characteristic of the human nervous system is that it is remarkably top heavy. The human nervous system con- sists of a large accumulation of cells at the top (the brain); a long, thin extension along the body axis (the spinal cord); and then still thinner extensions branching out into almost every part of the body. Why is the human nervous system arranged in this way? Doesn’t having so many neurons in a central location leave us vulnerable to injury? Wouldn’t it be safer to distribute the neurons evenly throughout the body, in a wide network?
In fact, some lower-level organisms do have a more even distribution of neurons throughout the body. For example, jellyfish and sea anemones have no centralized brain, but in- stead use a distributed nerve net to coordinate the slow, rhythmical contractions they use for movement and feeding. Nerve nets are the usual kind of nervous system for organ- isms with radial symmetry—that is, organisms with a top and bottom but no front, back, left, or right (FIGURE 2.2a). Although they do possess neurons, creatures like the jellyfish and sea anemones are literally spineless and brainless.
Unlike jellyfish, most animals have bilateral symmetry: the left side and the right side are almost mirror images of each other (FIGURE 2.2b). Bilateral animals have bodies built from a line of segments running from head to tail. You your- self are a segmented, bilateral organism. Your segments may not be as obvious as those of a centipede, but your ribs and the vertebrae of your spine are reminders of the underlying
FIGURE 2.2 Radial versus bilateral symmetry. (a) Organisms with radial symmetry, like sea anemones, may have distributed “nerve nets” but no centralized brain. (b) Organisms with bilateral symmetry, like dolphins, have a central nervous system running down their body segment. This central nervous system is enlarged at the head end into a brain.
(a)
(b)
02-Eagleman_Chap02.indd 39 02/11/15 3:16 pm
40 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 40 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
50–100 million years thereafter. We do know that almost all bilateral creatures, from the fruit fly to the mouse to the human, use a similar set of genes and signaling proteins (proteins that serve as chemical messengers between or within cells) to guide the development of the central nervous system. The signaling mechanisms are so similar that, in many cases, the gene of one organism (such as a mouse) can still perform its usual signaling function even when placed inside the nervous system of a dis- tantly related organism (such as a fruit fly) (FIGURE 2.4). This pro- vides evidence that the fundamental mechanisms for building a brain and spinal cord were already present before the ancestors of bilateral vertebrates (organisms with spines, like mice) and inver- tebrates (organisms without spines, like fruit flies) went down separate evolutionary paths, more than 550 million years ago.
It may seem strange to think of a mammal and an insect as having a common structure to their nervous systems. After all, the nervous system of a mammal runs along its back in a spinal cord, above the digestive tract, whereas the nervous system of an insect runs along its underside, below the digestive tract. How- ever, genetic studies of brain development suggest that the verte- brate nervous system may actually be an upside-down relative of the invertebrate nervous system (Gerhart, 2000) (FIGURE 2.5). For some reason, it appears as if the ancestor of all vertebrates evolved to live with its body flipped upside-down and then
FIGURE 2.3 Nervous systems of vertebrates and invertebrates. The nervous systems of vertebrates, as in this mouse embryo (a), have a spinal cord with segments and an expansion of the most anterior segments into a brain. The nervous systems of invertebrates, like this fruit fly Drosophila melanogaster (b), are also arranged into segments, with more complexity in the anterior segments (left).
(a) (b)
segmental organization of your body. The segmentation of your nervous system is an important feature, as will become more apparent later in this chapter.
For early bilateral animals, having a body of bilaterally symmetrical segments had a significant advantage: the animal could move its streamlined body swiftly through the water to search for prey or to avoid becoming prey. However, the con- trol requirements for a segmented organism with a front and a back are quite different from those of a jellyfish. The nervous system of bilateral organisms has two key features.
The first feature is the presence of local, centralized net- works within each body segment. In response to sensory input from the external world, simple circuits within each segment control the local muscles. In response to sensory input from the internal world of the body, similar circuits control the local internal organs. These circuits are capable of simple, localized functions like flexing a muscle when it is stretched.
The second feature is longitudinal transmission of infor- mation up and down the body axis between segments. Long connections up and down the length of the nervous system allow the activity of individual segments to be coordinated. For example, many organisms swim through the water by sending waves of alternating muscle contraction down the left and the right sides of the body. These waves of contrac- tion cause the body to undulate back and forth, propelling the organism forward through the water. Longitudinal con- nections and local, segmental circuits allow the organism to move itself through its environment efficiently.
But how did an early bilaterally symmetrical animal steer itself? The animal’s nervous system needed sensory inputs to tell it which way to go. Since the animal had a front end, it made sense to place the sensory equipment there (e.g., recep- tors for smell, taste, vibrations, light, and electrical currents). The front segments of the body then needed extra circuitry to deal with all of this extra input. Even more circuitry was needed to use this input to alter the ongoing activity of the circuits further back along the spinal cord, so that the animal could use this sensory input to guide what its body was doing. Additional circuitry was needed to drive the local muscula- ture of the mouth for feeding or the head for exploration. A ll these functions, and more, required extra neurons, and as these neurons were added in the course of evolution, the front end of the spinal cord began to bulge and expand into a top-heav y structure: the beginnings of a brain.
The Common Features of Every Central Nervous System Neuroscientists are still debating exactly how the evolution of a central nervous system took place or even how many times it took place. It is possible that both neurons and brains arose inde- pendently in several different lines of living creatures at different times (FIGURE 2.3). Specialized cells resembling neurons probably appeared more than 600 million years ago, and organized cir- cuits and structures made of neurons probably arose in the
FIGURE 2.4 The gene of one organism may still perform its signaling function even when placed inside the nervous system of a distantly related organism. (a) A fruit fly eye can be seen growing near its wing (left arrow), induced by inserting the fly’s Pax-6 gene into the local tissue. (b) An eye can also be induced to grow by the Pax-6 gene from a mouse, although fly and mouse diverged half a billion years ago.
(b)(a)
02-Eagleman_Chap02.indd 40 02/11/15 3:16 pm
An Overview of the Nervous System 41
# 158305 Cust: OUP Au: Eagleman Pg. No. 41 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
compensated by twisting just its foremost segments around 180 degrees, so that its mouth could still face downward toward the ground. The result for its human descendants is a brain whose uppermost parts are crossed over: a left hemisphere that con- nects with the right side of the body and a right hemisphere that connects with the left side of the body. All vertebrate brains, from those of hagfish to humans, are crossed over in this manner.
Vertebrates also share many other common features. They all contain a central nervous system consisting of the brain and the spinal cord. The spinal cord has input and output con- nections to the rest of the body via the peripheral nervous system. The peripheral nervous system connects not only to the skin and muscles, but also to the internal organs of the body.
The developing vertebrate brain itself contains three main bulges or zones of expansion: the forebrain (prosencephalon), the midbrain (mesencephalon), and the hindbrain (rhomb- encephalon). Even in the human brain, which grows quite complex at full development, these three fundamental bulges are clearly visible in a developing embryo at about 4 weeks. After this point, further subdivisions occur: the forebrain di- vides into the telencephalon and diencephalon, and the hind- brain divides into the metencephalon and myelencephalon. These structures, in turn, become further subdivided into the many structures of the adult brain, as shown in FIGURE 2.6 and as we will describe later in this chapter.
FIGURE 2.5 (LEFT) Vertebrate versus invertebrate nervous systems. The positions of the nervous system, gut, and circulatory system of vertebrates (a) are upside-down compared to that of invertebrates (b), as is the pattern of expression of major genes guiding body development.
FIGURE 2.6 (BELOW) Divisions within the developing vertebrate brain. The forebrain, midbrain, and hindbrain subdivide into all the substructures of the adult brain.
Nervous system
Gut (digestive system)
Circulatory system(a) Vertebrate
(b) Invertebrate
Cross section
Cross section
Prosencephalon (forebrain)
Telencephalon Cerebral hemispheres (cerebral cortex, subcortical white matter, basal ganglia, basal forebrain nuclei)
Thalamus Hypothalamus
Cerebral peduncles Midbrain tectum Midbrain tegmentum
Pons Cerebellum
Medulla
Diencephalon
Mesencephalon
Metencephalon
Myelencephalon
Spinal cord
Mesencephalon (midbrain)
Rhombencephalon (hindbrain)
Cerebral hemisphere
Diencephalon
Mesencephalon
Metencephalon
Myelencephalon
Telencephalon
Thalamus
Hypothalamus
Midbrain Brainstem Pons
Medulla
Cerebellum
Spinal cord
Spinal cord
(a) Five-vesicle stage (b) Embryo, lateral view (c) Adult
02-Eagleman_Chap02.indd 41 02/11/15 3:16 pm
42 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 42 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
functions of the simplest circuits, which lie in the peripheral nervous system and spinal cord. Then, gradually, we will move higher in the nervous system, seeing how the brain adds on layer after layer of modulation and refinement to the preexist- ing circuitry. We will conclude by discussing the system in the brain that unites the internal and external environments.
Getting Oriented in the Brain W hen describing locations in the nervous system, neurosci- entists use special terminology for anatomical directions (FIGURE 2.7). This terminology can be applied not only in humans, but also in any bilaterally symmetrical organism with a front, back, top, bottom, and sides. Rostral means toward the mouth, or the front end; the word comes from the Latin rostrum, or “ beak.” Caudal means toward the tail end; again, the word comes from the Latin caudum, or “tail.” Dorsal means toward the top (or back), from the Latin dorsum, or “ back.” Ventral means toward the belly, or bottom end, from the Latin venter, or “ belly.”
Anterior and posterior also mean toward the front or the back, respectively; superior and inferior mean toward the top or the bottom, respectively. Medial means toward the middle; lateral means toward the side. Ipsilateral means
Neuroanatomists have devised a detailed nomenclature to describe the vast number of structures in the human ner- vous system. It takes time and practice to become familiar with this nomenclature. However, a few basic principles will help you to understand the functions, rather than just the names, of brain areas.
First, remember that no neuron is an island. A ll neurons connect to other neurons through circuits, and almost all of these circuits are reciprocal: when neuron A sends output to neuron B, the odds are that neuron B also sends output back to neuron A. Second, remember that the role of a neuron in the nervous system as a whole depends largely on the neu- ron’s inputs and outputs. Knowing who sends input to a neuron and who receives its output can tell you a lot about what that neuron’s role is in the nervous system as a whole. Third, remember that when the brain refines one of its func- tions over evolutionary time scales, it often does so by insert- ing an additional layer of neurons between the existing inputs and outputs. These additional layers can help to mod- ulate the existing circuit, steering its activity more finely, in context with the circumstances at hand.
As we survey the nervous system, you will see how this process of modulation occurs and how it adds ever-increasing levels of complexity to the resulting behavior in an organism. To keep things simple, we will start with the structure and
(a)
(b)
Rostral
Caudal
Sagittal
Axial
Ventral
Dorsal
Ventral
Ventral
Dorsal
Dorsal
Coronal
FIGURE 2.7 Terminology for anatomical directions. (a) The basic anatomical directions in the brain. (b) The cardinal anatomical planes: sagittal (top), axial (middle), and coronal (bottom), from a standard human brain atlas, the Montreal Neurological Institute-152 atlas.
02-Eagleman_Chap02.indd 42 02/11/15 3:16 pm
The Peripheral Nervous System 43
# 158305 Cust: OUP Au: Eagleman Pg. No. 43 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
and blood vessels. They are sensitive not only to mechanical stresses such as pain or injury, but also to inflammation, fa- tigue, and temperature. Depending on the organ to which they provide input, they may also be sensitive to pressure, hor- monal or biochemical stimuli, or the local concentration of chemicals that are specific to each organ’s function. Sensory neurons have different names depending on what stimulates them: mechanoreceptors for physical movement, nociceptors for injury, thermoreceptors for temperature, baroreceptors for pressure, chemoreceptors for specific chemicals, and so on.
Motor neurons extend to the muscles of the body, making contact at a specialized structure called the neuromuscular junction. Electrical activity in the motor neuron causes a re- lease of signaling chemicals called neurotransmitters at the neuromuscular junction, and this in turn stimulates the muscle fiber to contract. (You will learn more about this pro- cess in Chapter 7 and more about neurotransmitters in Chapter 3.) Low levels of continuous (or “tonic”) motor neuron activity exist even at rest, and this activity produces a mild tension known as muscle tone. Higher levels of activity cause a vigorous contraction, which causes a body movement.
The body has two major compartments: the soma, in- cluding muscles, skin, and bones, and the viscera, containing the internal organs (FIGURE 2.8a). Other output neurons send signals to the visceral organs of the body. These visceral output signals regulate the activities of the body’s internal world: heart rate, respiration, blood pressure, temperature regulation, movements of the stomach and intestinal tract, secretion of digestive enzymes, voiding of the bladder and bowels, and sexual organ functions.
The division between the external and the internal world is reflected in the nervous system itself. The peripheral nervous system has two components: the somatic nervous system and the autonomic nervous system. The peripheral nervous system includes four kinds of neurons for input and output to these compartments: somatic afferent or somatosensory neu- rons (input), somatic efferent or motor neurons (output), vis- ceral afferent or visceral sensory neurons (input), and visceral efferent or autonomic neurons (output) (FIGURE 2.8b).
The somatic nervous system includes the sensory inputs and motor outputs for guiding voluntary body movements in the external world. W hen you raise your arm, kick a ball, or withdraw your hand from a hot plate, you are using the so- matic nervous system. In contrast, the autonomic nervous system regulates the body’s internal world. This process usu- ally goes on automatically (hence the label “autonomic”). W hen you digest your lunch, or your heart rate speeds up at the sight of an angry dog, or the hot sun causes your skin to sweat, your body’s autonomic nervous system is at work.
The autonomic nervous system is itself divided into two subsystems with opposite functions: the sympathetic ner- vous system and the parasympathetic nervous system (FIGURE 2.9). These subsystems allow the internal world of the body to operate in two basic modes.
The sympathetic nervous system puts the body in the mode of reacting to threats or opportunities in the external
“on the same side”; contralateral means “on the opposite side.” On a body extension such as a limb, distal means toward the far (“distant”) end of the limb, whereas proximal (from the Latin proximus, “nearest”) means toward the point where the limb attaches to the body.
Using these terms with respect to the human body can sometimes get confusing because although we keep our body axis vertical, our head still faces forward. Hence, in a human being, rostral can mean “anterior” for structures in the head (since “toward the mouth” will be the same as “toward the front”) but superior for structures in the rest of the body (since “toward the mouth” will be the same as “toward the top”). Likewise, dorsal can mean “above” for structures in the head but “behind” for structures in the rest of the body. If this is confusing, you may find it easier to imagine the human body on all fours when thinking about anatomical directions.
Neuroscientists often view the nervous system in planes, or slices: a microscope section, a computerized tomography scan, or an MR I series (see Chapter 1 for a more detailed de- scription of computerized tomography and MR I). Hence, they have developed a set of labels to indicate how the slices are oriented in space. An axial slice divides the body along its long axis, into rostral and caudal. A sagittal slice divides the body into left and right. The name comes from the Latin sagit- tus or “arrow,” as if an imaginary arrow down the spine would pass through this plane. A midsagittal slice is a slice through the exact midline of the body or nervous system. A frontal or coronal slice divides the body into dorsal and ventral. The name comes from the Latin for “crown,” as if someone were to place an imaginary crown on the head along this plane.
Now that we have been oriented in the brain and have learned the terminology for anatomical directions, let’s begin our exploration of the brain from its outermost ex- tremities, in the peripheral nervous system.
The Peripheral Nervous System As previously stated, the peripheral nervous system con- nects the spinal cord and the rest of the body. Let’s examine some aspects of its structure and function.
Separate Systems for the Inner and Outer Environments Sensory neurons have receptors in the skin, muscles, and joints, and through these they convey a multitude of different kinds of sensory input to the body: touch, vibration, pain, tem- perature, fatigue, itch, stretch, and position. Other sensory nerves extend into the visceral organs of the body: heart, lungs, stomach, intestines, pancreas, kidneys, bladder, uterus,
02-Eagleman_Chap02.indd 43 02/11/15 3:16 pm
44 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 44 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.8 The body and the peripheral nervous system. (a) The body has two major compartments: the soma and the viscera. (b) The peripheral nervous system includes four kinds of neurons for input and output to these compartments: somatosensory neurons (input), motor neurons (output), visceral sensory neurons (input), and autonomic neurons (output).
Spinal cord
(a) (b)
Muscle
Bone
Skin
Liver
Stomach
Pancreas
Kidney
Intestines
Bladder
Lungs
Heart
Soma: Viscera:
Somatosensory neurons (input) Motor neurons (output)
Visceral sensory neurons (input) Visceral (autonomic) motor neurons (output)
world: feeding, fighting, fleeing, or sexual activity. In this mode, the heartbeat quickens, respiration increases, blood pressure increases, and circulation shifts from the digestive organs to the muscles, while the movements of the digestive tract itself slow down or come to a halt. This “fight-or-flight” response system prepares the body to deal with urgent mat- ters in the external world.
In the absence of urgent matters, priorities shift from fight- or-flight to “rest-and-regenerate”—and this latter mode is con- trolled by the parasympathetic nervous system. In this state, the heart rate slows, respiration decreases, blood pressure falls, muscle tone relaxes, and blood flow shifts to the stomach and digestive organs. The movements of the digestive tract in- crease as the body rebuilds its stores of energy, protein, and other nutrients. The next time you and a friend or family member are suffering from having eaten too much food in one sitting, you can tell each other, “I parasympathize.”
A Nervous System with Segmental Organization The peripheral nervous system is not an evenly distributed nerve net like that of a jellyfish. As we discussed earlier, the human nervous system has a segmental organization (FIGURE 2.10). The segments are easiest to see near the spinal column. Here,
the body segments are apparent in the skeleton itself: the line of vertebrae that begins with the cervical spine (within the neck), continues down the thoracic spine (within the rib- cage, another segmented structure), continues farther down the lumbar spine (between the ribcage and the pelvis), and ends in humans with the sacral spine (whose segments are fused together into the bony, triangular “sacrum” that forms the back of the pelvis).
The peripheral nerve roots emerge from the spinal cord on either side, near the junction of each vertebra with its neighbor. Hence, every segment of the spinal cord has its own set of peripheral nerve roots on the left and on the right. Near the spinal cord, inputs and outputs are kept separate. A ll sensory input, somatic and visceral, enters the spinal cord through the dorsal nerve root at the back of the spinal cord. A ll motor output, somatic and autonomic, exits the spinal cord through the ventral nerve root at the front of the spinal cord.
The segmental organization of the somatic nervous system can also be observed on the outside of the body. Each
02-Eagleman_Chap02.indd 44 02/11/15 3:16 pm
The Peripheral Nervous System 45
# 158305 Cust: OUP Au: Eagleman Pg. No. 45 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.9 The autonomic nervous system directs the activity of the visceral organs and has two divisions. The sympathetic nervous system directs activities for fight-or-flight responses, whereas the parasympathetic nervous system directs activities for rest and regeneration.
Sympathetic nervous system Parasympathetic nervous system
Eye
Salavary glands
Trachea
Bronchi
Parasympathetic ganglia
Heart
Liver
Gallbladder
Adrenal gland
Kidney
Stomach
Preganglonic fibers
Paravertebral sympathetic ganglia
Postganglonic fibers
Celiac ganglion
Superior mesenteric ganglion
Inferior mesenteric ganglion
Alimentary tract
Urinary bladder
pair of sensory nerve roots handles input from a narrow stripe on the body surface. The stripes are arranged in a series, from head to tail end, as if the skin were cut up into narrow sections. These stripes are therefore called derma- tomes, from the Greek words for “skin” and “to cut.” Each dermatome corresponds to a different spinal segment and is numbered accordingly—for example, the dermatome of the fifth segment of the cervical spinal cord is known as C5 (FIGURE 2.11a). If the spinal segment or its nerve roots are in- jured, sensation in the dermatome may be lost. A physician can often determine the site of a spinal injury by carefully testing for areas of sensory loss on the body surface. As one way of visualizing a dermatome, you may be familiar with
shingles, an illness that results from a dormant form of the Varicella zoster (chickenpox) virus reactivating itself in one of the nerve roots of a person’s body. This causes a painful, blistering skin rash to appear along the dermatome of the nerve root where the virus has reactivated—thus making the dermatome visually apparent (FIGURE 2.11b).
The motor side of the somatic nervous system is also or- ganized into segments. In this case, the segmental “stripes” lie within the musculature rather than the skin and are there- fore called myotomes rather than dermatomes. However, they, too, are arranged in a series from head to tail. Spinal injuries can also affect output to the myotomes. For exam- ple, an injury to the thoracic spine can block output from the brain to the leg muscles, causing lower limb paralysis (para- plegia). Higher up, an injury to the neck can block output to both the lower and the upper limbs, producing paralysis of all four limbs (quadriplegia). The segmental level of a spinal injury will determine which parts of the body are disabled and which are spared.
02-Eagleman_Chap02.indd 45 02/11/15 3:16 pm
46 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 46 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Trigeminal
Cervical
Thoracic
Lumbar
Sacral
C2 C3
C4 C5C6C7C8
T1
T12 S1
T11 T10
T9 T8
T7 T6
T5 T4
T3 T2
S2
S1
L5
L4
L2
L3
L1
S2
S2 L4
L5
L3 L2
L1
T1 C7
C6
C8
C5 C6
C7
C8
C7
(a) (b)
FIGURE 2.11 (a) Sensory dermatomes of the human body and (b) a shingles rash affecting a thoracic dermatome.
FIGURE 2.10 Segmental organization of the peripheral nervous system.
C1 C2 C3 C4 C5 C6 C7 C8 T1
T2 T3
T4 T5 T6 T7 T8 T9 T10
T11
T12
S1 S2
S3 S4
S5
Cervical spine
Lumbar spine
Sacral spine
Thoracic spine
Somatosensory (input) Motor (output) Visceral sensory (input) Visceral (autonomic) motor (output)
Dorsal nerve root Spinal cord
Ventral nerve root
Autonomic ganglionL1
L2
L3
L4
L5
02-Eagleman_Chap02.indd 46 02/11/15 3:16 pm
The Spinal Cord 47
# 158305 Cust: OUP Au: Eagleman Pg. No. 47 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
paths between the spinal cord and their endpoints, so that the organization can appear scrambled at intermediate points along the way.
The Spinal Cord The spinal cord is the meeting place for the inputs and out- puts of the peripheral nervous system. Linking sensation to appropriate action is the fundamental method by which ner- vous systems help an organism to survive and reproduce. In the spinal cord, we find rudimentary circuits that use sen- sory input to guide motor output. W hat do these circuits look like, what can they do, and how can they do it? In this section, we’ll address each of these questions in turn.
Circuits within a Segment: Spinal Reflexes If we take a cross-section through a spinal cord, we see a small central canal, surrounded by a butterfly-shaped struc- ture made of gray matter, which is itself surrounded by an oval of white matter. As throughout the rest of the brain, the central gray matter is home to the cell bodies of neurons and
The neat, segmental pattern of the myotomes and derma- tomes is not perfectly preserved at all points in the peripheral nervous system. Sensory and motor neurons take compli- cated routes from the spinal cord to their final destinations, joining and rejoining neurons from other spinal levels as they reshuffle themselves into peripheral nerve bundles.
The autonomic nervous system is also organized into segments. In fact, the sympathetic and parasympathetic sys- tems exist in completely separate segmental regions. A ll sympathetic outputs come exclusively from the middle levels of the spinal cord: the thoracic segments and the neighbor- ing, uppermost lumbar segments. A ll parasympathetic out- puts come from either the tail-end segments of the sacral spinal cord or the head-end segments that lie above the spinal cord altogether, in the brainstem.
If this seems complex, just keep in mind some basic prin- ciples. The peripheral nervous system has a somatic system for controlling the movement of the “soma,” or body, through the external environment and an autonomic system for con- trolling the “automatic” responses of the internal organs. Both systems have sensory inputs and motor outputs. The autonomic outputs fall into separate subsystems for switch- ing between the body’s fight-or-flight (sympathetic) and rest-and-regenerate (parasympathetic) modes. Both systems have a segmental organization of input and output pathways at their point of entry or exit to the spinal cord. However, the input and output pathways often take complicated, tortuous
CASE STUDY: Christopher Reeve, 1952–2004 The actor Christopher Reeve was known and admired by a generation of moviegoers for playing the role of the fictional hero Superman in the 1970s and 1980s. The character he portrayed was endowed with hercu- lean strength, an almost total invul- nerability to injury, and the ability to fly. Tragically, Reeve himself became almost completely paralyzed after being thrown headfirst from a horse during an equestrian competition in 1995. Although his helmet prevented direct brain injury, the force of the impact destroyed the first and second vertebrae of his neck, crush- ing the upper cervical spinal cord. He was left unable to move his arms
or legs and unable to even breathe without mechanical assistance. For the rest of his life, he was restricted to a powered wheelchair, which he steered by puffing air through a strawlike device. The wheelchair had a built-in ventilator to supply his lungs with air.
Undaunted by his injury, Reeve became a prominent activist for spinal cord research. Over the next nine years he traveled the world, making speeches and raising mil- lions of dollars to help fund the search for a cure and improve the quality of life for those with spinal injuries. He urged the U.S. govern- ment to support research on using
stem cells (undifferentiated cells that have the ability to differentiate into several cell types) to regener- ate injured neurons. Throughout these years, he maintained an in- tense regimen of physical therapy and ultimately did regain some sen- sory and motor function. However, he also battled a series of infections from the bed sores caused by his immobility. In 2004, he died of heart failure at the age of 52. His enduring optimism remains an inspiration to those who live with neurological in- juries and illness. Today, the Chris- topher and Dana Reeve Foundation continues to fund research toward a cure for spinal cord injury.
02-Eagleman_Chap02.indd 47 02/11/15 3:16 pm
48 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 48 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
and visceral (FIGURE 2.12b). Somatic sensory neurons take input mostly from the skin, skeletal muscles, and joints. This input is useful mainly for guiding body movements with the skeletal muscles: position, stretch, and touch. By contrast, visceral sen- sory neurons take input from many other tissues as well, includ- ing the internal organs. Such input is useful mainly for regulating the internal state of the organism: temperature, pain, inflamma- tion, fatigue, and so on. Sensory neurons are also known as so- matic or visceral afferents, the word “afferent” meaning that their information is “carried toward” the central nervous system.
The motor neurons of the spinal cord also lie in separate somatic and visceral columns. Somatic motor neurons send output signals to stimulate the muscles lying in the local myotome. Visceral motor neurons are the output neurons of the sympathetic and parasympathetic nervous systems, sending control signals to the body’s internal organs. Motor neurons are also known as somatic or visceral efferents, the word “efferent” indicating that they “carry away” informa- tion from the central nervous system.
Now that we have our afferents and efferents in place, what can we do with them? The simplest kind of circuit for connect- ing inputs to outputs is called a reflex arc. In this kind of circuit, a sensory neuron makes an excitatory connection to a motor neuron so that when the sensory neuron is stimulated, it acti- vates the motor neuron in return. This kind of reflex arc is useful in helping muscles compensate for additional load. If a sensory neuron detects that the muscle is being stretched, it stimulates the appropriate motor neuron to contract the muscle.
The familiar “knee-jerk ” reflex (properly called the pa- tellar tendon reflex) is a result of suddenly stretching the body of the quadriceps muscle after tapping the attached tendon with the reflex hammer (FIGURE 2.13). The quadriceps makes a powerful, automatic contraction in response to this unexpected extra load. Reflex arcs are just as important for coordinating the activity of the sympathetic and parasympa- thetic nervous systems: raising the hair follicles in response to cold, producing tears in response to eye irritation, and contracting the blood vessels when standing up so as not to faint from loss of consciousness.
A reflex involving a direct connection between a sensory and a motor neuron is sometimes called a monosynaptic reflex, since the entire circuit involves only one synapse, or connection between neurons (we’ll learn more about syn- apses in the next chapter). However, these kinds of circuits are rare. Most reflexes are polysynaptic reflexes—that is, they involve more than one synapse, because an interneu- ron lies between the incoming sensory neuron and the out- going motor neuron in the circuit.
Compared with monosynaptic reflexes, polysynaptic re- flexes allow for more flexibility in the response. For example, imagine you need to contract the biceps muscle of your arm. The problem is that this will stretch the triceps muscle on the other side of the arm. As we saw above, the triceps will auto- matically compensate by fighting against the stretch. With- out some way of turning off the triceps stretch reflex, you cannot bend your arm—effectively, you are frozen in place.
their local connections, whereas the surrounding white matter is made up of the electrically insulated, long-distance connections between neurons. The overall size of the cord depends on the body segment. If a body segment contains a limb or part of a limb, it will need a larger cord with more gray matter to handle the extra sensory and motor informa- tion and more white matter to handle the extra communica- tion with the distant neurons in the brain.
The neurons of the gray matter are stacked in layers, or laminae, from ventral to dorsal. As we saw in the last sec- tion, sensory input enters the cord from the dorsal side, whereas motor output exits the cord from the ventral side (FIGURE 2.12a). Hence, as you might expect, the neurons in the dorsal layers are mostly sensory neurons, whereas the neurons in the ventral layers (also called the ventral horns) are mostly motor neurons. The cell bodies of peripheral sen- sory neurons live just outside the spinal cord, in the dorsal root ganglion. Incoming peripheral signals pass these cell bodies on the way to the spinal cord.
The sensory neurons of the upper gray matter layers lie in two separate columns (perpendicular to the laminae): somatic
FIGURE 2.12 Transverse section of a spinal cord. (a) Laminae and sensory input and motor output through the dorsal and ventral nerve roots as well as (b) the zones for somatic and visceral sensory (upper) and visceral and somatic motor (lower) neuron cell bodies in the spinal cord gray matter.
Somatosensory (input)
Somatic motor (output) Visceral sensory (input)
Visceral motor (output)
Dorsal nerve root
White matter
I II
III
IV
V
VI
VII
VIII
IX
X
(a) (b)
Ventral nerve root
Gray matter
Central canal
Marginal zone
Gelatinous substance
Nucleus proprius
Lateral motor neurons
Medial motor neuronsVentral
Dorsal
02-Eagleman_Chap02.indd 48 02/11/15 3:16 pm
The Spinal Cord 49
# 158305 Cust: OUP Au: Eagleman Pg. No. 49 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Brown concluded that the main functional unit of the nervous system was not the reflex arc, but a central pattern generator capable of generating movement activity sponta- neously (FIGURE 2.14). Input from peripheral sensory neurons of higher brain centers could then adjust these intrinsic pat- terns of activity as needed. Over the past century, neurosci- entists have discovered many different examples of central pattern generators in many species (Barlow, 2009; Frigon, 2012; Konishi, 2010). These circuits drive rhythmical pat- terns of movement in many different parts of the body: arms, legs, wings, fins, tails, jaws, breathing apparatuses, and vis- ceral organs such as the stomach or intestinal tract.
The basic circuitry of a central pattern generator is sur- prisingly simple. As in a reflex arc, output comes from a motor neuron that excites the muscle to be moved. The motor neuron is driven by a nearby excitatory interneuron (a neuron that transmits excitatory signals between other neurons), which is spontaneously active even without input. The excit- atory interneuron becomes fatigued, and its activity gradu- ally wanes over time. A slightly weaker inhibitory interneuron connects to both of these neurons. As the excitatory inter- neuron becomes fatigued, the inhibitory interneuron eventu- ally becomes strong enough to shut down the activity of the other neurons, halting the muscle movement. This gives the excitatory interneuron time to recover until it eventually overcomes the inhibition and begins to fire once again.
If we place an identical circuit on the other side of the spinal cord, we can have a pattern of oscillating motor ac- tivity on each side of the body. Finally, if we place a pair of inhibitor y interneurons reaching across the midline to the opposite circuit, we can ensure that when one circuit is on, the other is off, and vice versa. This k ind of pattern will pro- duce alternating movements, such as walk ing. A lterna- tively, we can use excitator y interneurons, so that both circuits will turn on and off at the same time. This k ind of pattern will produce synchronized movements, such as hopping. We will learn more about central pattern genera- tors in Chapter 7.
But what if we add an inhibitory interneuron (a neuron that transmits inhibitory signals between other neurons) that connects the biceps motor neuron to the triceps motor neuron? Now the triceps motor neuron has information about the contraction of the biceps and can use this informa- tion to override the stretch reflex. You are no longer frozen: you can now bend your arm and wipe your brow in relief.
Visceral and somatic reflexes within a single segment can perform many local functions. However, to execute more complicated actions, we must coordinate activity across multiple body segments. Let’s now take a look at the more elaborate circuits that extend across segments and see what feats they are able to accomplish.
Complex Circuits across Segments: Central Pattern Generators The simple circuits of reflex arcs were among the first to be studied in the nervous system. During the late 19th and early 20th centuries, the British physiologist Sir Charles Sherrington examined how the neural pathways of the spinal cord integrated sensory information to help guide basic forms of movement. He proposed that more complex motor behaviors, such as locomotion (movement from one place to another), could be generated by chains of simpler reflex ac- tions, all driven by incoming sensory input (Burke, 2007).
But was sensory input truly necessary to drive the pro- cess, or can the nervous system generate the necessary move- ments for locomotion all on its own? A few years later, Sherrington’s student, T. Graham Brown, found that anes- thetized cats could still continue to make stepping move- ments, even when their spinal cords were completely deprived of all peripheral sensory input. In fact, even when isolated from the brain itself, the spinal cord alone could still drive the stepping movements without sensory input (Brown & Sherrington, 1911; Stuart & Hultborn, 2008).
FIGURE 2.13 The patellar tendon reflex. Tapping the tendon stretches the quadriceps muscle, sending a sensory signal to the spinal cord and triggering a reflex contraction of the same muscle, along with a relaxation of the opposing muscle.
Afferent (sensory) fiber
Motor axon (active)
Motor axon (inhibited)
Interneuron
Dorsal nerve root
Quadriceps (extensors)
Hamstrings (opposing flexors)
Patellar tendon
Ventral nerve root
02-Eagleman_Chap02.indd 49 02/11/15 3:16 pm
50 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 50 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
THE BIGGER PICTURE: In Search of a Cure for Spinal Cord Injury The Edwin Smith Surgical Papyrus is one of the world’s most ancient sur- viving medical texts. Dating from more than 3,500 years ago, it de- scribes with remarkable detail and accuracy the diagnosis and treat- ment for 48 cases of traumatic bodily injury. Here are its instructions to a physician faced with a patient who has a damaged spinal cord: Thou shouldst say concerning him: “[This patient has] a dislocation in a vertebra of his neck, while he is unconscious of his two legs and his two arms, and his urine dribbles. An ailment not to be treated.”
Unfortunately, even in today’s modern age, the debilitating effects of spinal cord injury remain perma- nent in most cases. Neither the neu- rons nor the white matter tracts of the cord regenerate after they are damaged. On the contrary, in the
days and weeks postinjury, the pa- tient’s physical condition worsens. The injured area suffers from in- flammation, cell death, and the toxic effects of released cell contents from dying neurons. Ultimately, an empty cavity forms, surrounded by a barricade-like scar of glial cells (cells that support neurons), many times the size of the original injury. These physical obstacles prevent neurons from regrowing connec- tions throughout the damaged area.
One potential treatment cur- rently under investigation is to im- plant neural stem cells into the spinal cord shortly after injury (Li & Lepski, 2013; Mothe, Tam, Zahir, Tator, & Shoichet, 2013). These cells are not yet fully developed into their final forms, and they are there- fore uncommonly versatile. Neural stem cells typically proceed to
differentiate into neurons or into cells that support neurons (i.e., glial cells—about which you will learn in Chapter 3). Experimental studies show that neural stem cells may be able to grow new neural pathways to replace those that have been lost (Liu et al., 2013; Saadai et al., 2013). However, the physical barriers of the scar and cavity from the destroyed neural circuits must still be over- come somehow. Potential risks also surround the use of neural stem cells. The embryonic stem cells may activate the patient’s immune system (Xu et al., 2010), divide excessively, form pain pathways rather than motor pathways (Macias et al., 2006), or even become a tumor (Kuroda, Yasuda, & Sato, 2013). Many ethical concerns surround the use of em- bryonic stem cell therapy. The po- tential for the development of safe
Right electrode
Motor neuron Excitatory interneuron Inhibitory interneuron
Left motor neuron
Left motor neuron
Right motor neuron
Right motor neuron
Right electrode
Left electrode
Dorsal
Right Left
Ventral
Left electrode
Swimming (lamprey) Walking (cat)
(a) (b)
FIGURE 2.14 Central pattern generator circuits. These circuits in the spinal cord fire in alternating left–right patterns to drive locomotion in animals as diverse as lampreys swimming (a) or cats walking (b).
02-Eagleman_Chap02.indd 50 02/11/15 3:16 pm
The Brainstem 51
# 158305 Cust: OUP Au: Eagleman Pg. No. 51 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
and effective neural stem cell ther- apy holds great promise.
In the meantime, treatment of spinal cord trauma focuses on trying to reduce the inflammation
and cell death in the hours and days after injury using medications such as corticosteroids, which have pow- erful anti-inflammatory and immu- nosuppressive effects. This may
help to limit the damage, but does not heal the neural circuits already destroyed. Finding a cure for spinal cord injury remains a Holy Grail of 21st-century neuroscience.
FIGURE 2.15 Structures of the brainstem, seen in (a) lateral view and (b) ventral view.
(a) (b)
Optic tract
Optic tract
Optic chiasm
Cerebral peduncle
Cerebral peduncle
Pyramid decussation
Pyramid
Inferior olive
Middle cerebellar peduncle
Oculomotor nerve (CN III)
Lateral geniculate body
Anterior
Posterior
LeftRight
Superior
Inferior
PosteriorAnterior
Medial geniculate body
Superior colliculus
Inferior colliculus
Trochlear nerve (CN IV)
Trigeminal nerve (CN V)Cerebellar peduncles: Superior Middle Inferior
Vagus nerve (CN X)
Spinal accessory nerve (CN XI)
Hypoglossal nerve (CN XII)
Abducens nerve (CN VI)
Facial nerve (CN VII)
Trigeminal nerve (CN V)
Abducens nerve (CN VI)
Facial nerve (CN VII) Vestibulocochlear nerve (CN VIII)
Glossopharyngeal nerve (CN IX)
Pyramid
Dorsal (posterior) columns: Fasciculus gracilis
Fasciculus cuneatus
Olive
Optic nerve (CN II)
Optic nerve (CN II)
Thalamus
Midbrain
Pons
Medulla
Spinal cord
The Brainstem The brainstem, which is the most posterior region of the brain, acts as a point of communication between the spinal cord and the most anterior structures of the nervous system (FIGURE 2.15). It is composed of three structures: the medulla oblongata, pons, and midbrain. The brainstem’s most caudal structure is the medulla oblongata. Ahead of it, the brain- stem becomes riddled with the additional white matter tracts of the pons or “ bridge,” which provides connections to the elaborate circuitry of the cerebellum (“little brain”) at the same level. Ahead of these two structures lies the midbrain (or mesencephalon) and beyond it the rest of the brain.
Extending out from the brainstem are the cranial nerves. Let’s take a survey of each one of these structures in turn.
Medulla Oblongata and Pons The medulla oblongata and pons form the hindbrain. The medulla oblongata regulates involuntary functions that are essential to life, including breathing, heart rate, and blood pressure. The pons relays signals between the cerebellum and the cerebrum (the cerebrum is the anteriormost struc- ture of the central nervous system, consisting of the cerebral cortex, basal ganglia, hippocampus, and amygdala. It origi- nates from the telencephalon of the developing embryo). The
02-Eagleman_Chap02.indd 51 02/11/15 3:16 pm
52 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 52 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
these brainstem nuclei determine the direction and speed of motion, they must pass this information to the motor nuclei controlling each eye’s movements. This in itself is complicated because six different muscles and three different nuclei con- trol eye movement. These movements are not even symmetri- cal: looking to the left means pulling your right eye toward the nose but your left eye away from the nose. Many apparently simple behaviors, like keeping the eyes steady during move- ment, require a complex circuitry behind the scenes.
Hindbrain circuits also act as central pattern generators for the rhythmical movements of the head and upper body. Chewing, swallowing, yawning, sucking, coughing, sneez- ing, and hiccupping are examples of rhythmical movements generated by the brainstem. However, the most important central patterns are the movements of breathing. In the me- dulla oblongata, at least two central pattern generators drive the rhythmical movements of breathing, and either one is sufficient to maintain respiration on its own. Visceral central pattern generators in the medulla oblongata also perform critical functions: the regulation of heart rate and blood pressure. With such heav y responsibilities, the medulla ob- longata is essential to survival. A lthough the body can endure many other kinds of brain injury, destruction of the medulla oblongata is swiftly fatal.
pons is involved in arousal, sleep, breathing, swallowing, bladder control, eye movement, facial expressions, hearing, equilibrium, and posture. In many ways, the hindbrain re- sembles the spinal cord in structure and function. For ex- ample, the hindbrain has incoming sensory neurons and outgoing motor neurons that form peripheral nerves. The hindbrain, too, has a central gray matter with different col- umns of neurons interacting with the inside and outside worlds: somatic sensory, visceral sensory, visceral motor, and somatic motor neurons. The spinal cord and hindbrain also have some similar kinds of circuits such as reflex arcs.
However, the hindbrain has many important sensory fea- tures that the spinal cord lacks. Keep in mind that the head of a bilateral organism has many kinds of special sensory organs not found in the rest of the body. For example, it has light re- ceptors in the eye, sound receptors in the inner ear, odor recep- tors in the nasal passages, and taste receptors in the oral cavity. In addition, it has pain, temperature, vibration, position, and touch receptors of the head and inside the mouth and throat, as well as the balance-sensing vestibular organ of the inner ear. Many creatures have other sensory systems as well: the shark and the platypus have electroreceptors, pigeons have magne- toreceptors, rattlesnakes have infrared-detecting pit organs below the eyes, and fish have sensitive vibration detectors run- ning down the sides of their bodies. All these extra sensory inputs travel to the brainstem, which needs additional neural circuitry to handle them.
Likewise, the hindbrain has many important motor fea- tures that the spinal cord lacks. The tongue, mouth, neck, and head have different forms of movement than the rest of the body, so they need new kinds of circuitry. The eyes cannot form a stable image without expert control. They need an elab- orate musculature and an even more elaborate control system to keep themselves steady and on target as the body moves around. Also, let’s not forget the visceral side of motor control. The swallowing actions of the upper throat (or pharynx) need careful coordination, as do the vibratory movements of the vocal apparatus, or larynx, of land animals and the more elab- orate syrinx of singing birds. Water-breathing animals need to ventilate their gills. Air-breathing animals need to ventilate their lungs. Heartbeat, blood pressure, digestion, the voiding of the bladder and bowels, the functions of the sexual organs: all these visceral functions will also need central control to ensure that their activity is appropriate to the circumstances. For a bilateral animal, steering is complicated.
Brainstem nuclei handle these new sensory, motor, so- matic, and visceral functional requirements (FIGURE 2.16). Simple brainstem reflex arcs can handle simple, local re- sponses. For example, brushing the sensitive surface of the eye provokes a vigorous, protective blinking movement of the eyelid on the same side. However, most brainstem reflexes are much more complicated. For example, the vestibulo-ocular reflex keeps the eyes steady if the head is suddenly turned or moved. This relatively basic function requires many nuclei to work together. Detecting the sudden movement requires sen- sory nuclei to process input from the vestibular organ. As
FIGURE 2.16 Nuclei of the brainstem. (a) The neurons of the brainstem are arranged into nuclei, which themselves are arranged into columns. There are different columns for visceral and somatic sensory and motor functions. (b) Along each column, different nuclei serve different parts of the head and neck region via the cranial nerves. In this figure, on the left side, only the motor nuclei are shown. On the right side, only the sensory nuclei are shown.
Somatic motor nuclei
Somatic sensory nuclei
Visceral motor nuclei
Visceral sensory nuclei
Fourth ventricle
Section through brainstem
(b) (a)
02-Eagleman_Chap02.indd 52 02/11/15 3:16 pm
The Brainstem 53
# 158305 Cust: OUP Au: Eagleman Pg. No. 53 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
NEUROSCIENCE OF EVERYDAY LIFE: Why Do We Get the hiccups? Most central pattern generators of the brainstem perform functions es- sential to life: swallowing, breath- ing, and so on. There is one annoying exception, however: the hiccups. A hiccup involves closing the vocal folds of the glottis and then making a sudden inspiratory movement of the diaphragm. The movement is similar to a cough, except that in a cough the diaphragm makes a sudden expira- tory movement, hopefully blasting any obstructions out of the airway. The importance of a clear airway is obvious, but what could be the pur- pose of a hiccup?
In 2003, researchers at the Uni- versity of Calgary proposed that the answer may lie with our amphib- ian ancestors (Straus et al., 2003). These animals possessed simple lungs, but also had gills for respira- tion in water. When under water, many amphibians ventilate their gills by pumping water over them with rhythmic contractions of the pharynx. During this gill breathing, the glottis closes forcefully to pre- vent water from flooding the lungs. You may have seen the rhythmic pul- sations of a frog’s throat as it rests. These are remnants of movements
from their gill-ventilating youth as tadpoles, although the adult frog has lost its gills. Although we, too, lost our gills long ago, the central pattern generator for gill ventilation may have persisted in our medulla oblongata. Part of this central pat- tern generator is still useful for controlling suckling behaviors in in- fants—another function essential to life. Hiccups might therefore be the price we pay for being able to suckle as infants.
Midbrain The midbrain adds yet another layer of modulation and com- plexity to the circuits we have seen so far (FIGURE 2.17). Its local inputs and outputs come mostly from the eyes: visual signals from the retina, motor signals to control eye move- ments, light entry via the iris, and image focus via the lens. To make use of a visual image, the animal must determine where objects are in space.
A midbrain area called the superior colliculus (col- liculus means “ little hill,” based on its external appear- ance) is involved in locating visual stimuli in space and uses this information to direct complex movements, such as turning the eyes to point toward the target. This area can also use visual stimuli to guide movements of other parts including the head, the arm, the tongue (in frogs), or even the whole body. Just below the superior colliculus, the in- ferior colliculus performs parallel functions using audi- tor y rather than visual inputs. These areas are large, important players in motor control in many species such as fish, amphibians, reptiles, and birds. In mammals, many of the functions of the superior colliculus and the inferior colliculus have been transferred to the cerebral cortex, as we will see later in this chapter.
The midbrain also contains so-called command genera- tors capable of starting, stopping, and modulating the activ- ity of central pattern generators in the brainstem and spinal
cord. An example of a command generator is the midbrain locomotor region, a set of nuclei in the midbrain capable of initiating locomotor movements (movements from one place to another) whose nature depends on the species: swimming in aquatic vertebrates or walking in land vertebrates. Placing the command generator in the midbrain allows it to take input from many different sources: visual or auditory input, startle reflexes, and more anterior brain areas responsible for the kinds of behavior that require movement (such as explor- ing, seeking, and evading).
Some of these t y pes of complex behav ior arise from the circuitr y of a central midbrain area called the periaque- ductal gray matter. Here the neurons are again organized into a set of columns, as in the spinal cord. In this case, however, each column handles a different basic class of behav ior: not simple behav ior like coughing or yaw ning, but more complex behav ior like defense, aggression, or reproduction.
Three major categories of behavior are needed for sur- vival and reproduction. These are appetitive behaviors for finding and consuming essential nutrients, agonistic behav- iors for attack and defense against hostile organisms, and reproductive behaviors for courting, mating, and rearing off- spring. The midbrain coordinates these behaviors, each of which involves integrating many simpler components, both somatic and visceral. For example, stimulating one of the columns of the periaqueductal gray matter in a cat can elicit a stereotyped defensive reaction: flattening the ears,
02-Eagleman_Chap02.indd 53 02/11/15 3:16 pm
54 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 54 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.17 Structures and nuclei of the midbrain.
Superior colliculus
Inferior colliculus
Red nucleus
Reticular formation
Periaqueductal gray matter
Cerebral aqueduct
Reticular formation
Ventral tegmental area Locus
coeruleus
Substantia nigra
Dorsal raphe nuclei
narrowing the eyes, baring the teeth, striking out with the forepaw, growling with the vocal cords, and increasing the heart rate and blood pressure (Bhatt & Siegel, 2006). Stimu- lating other columns can elicit other behavioral patterns: the species-specific responses of courtship, or mating, or freez- ing, or urinating, or defecating (Drake et al., 2010; Kow, Brown, & Pfaff, 1994; Ohmura et al., 2010; Schimitel et al., 2012). In short, the periaqueductal gray is a sort of central pattern generator that coordinates not reflexes or move- ments, but other pattern generators themselves.
The midbrain coordinates many important activities not only in the hindbrain and spinal cord, but also in the forebrain. A diff use network of midbrain cells k nown as the midbrain reticular formation plays a central role in regulating states of consciousness. Depending on its activ- ity, the forebrain will settle into the alertness of the wak ing state or the unconsciousness of the sleeping state. A nother cell group called the locus coeruleus (or “ blue place” be- cause of its distinctive pigmentation) sends alerting sig- nals to the rest of the brain via a neurotransmitter called norepinephrine. Other key neurotransmitter systems are also headquartered in the midbrain. The substantia nigra (or “ black substance,” again because of its color) is the main source of the neurotransmitter dopamine, which plays key roles in movement, cognition, motivation, and reward. The substantia nigra is also part of the basal gan- glia, about which you will learn later in this chapter. The midbrain raphe nuclei, which lie along the seam between the two sides of the brainstem, are the main source of the neurotransmitter serotonin. Serotonin has diverse func- tions in mood, sleep, and social behavior. Each of these neurotransmitter systems is in itself a major topic, and we’ ll explore each neurotransmitter system in detail in a later chapter.
Most Cranial Nerves Emerge from the Brainstem Humans have 12 pairs of cranial nerves, which are some- times numbered with Roman numerals. The cranial nerves transmit sensory and motor information between the brain and the periphery, similar in some ways to the peripheral nerves that connect to the spinal cord. A ll of the cranial nerves emerge from the brainstem, except for cranial nerves I and II (which emerge from the cerebrum itself). (The cere- brum is the anteriormost structure of the central nervous system). FIGURE 2.18 shows each cranial nerve, along with its major function(s).
The Cerebellum As we have seen, the spinal cord and brainstem can produce basic forms of motor activity using reflex circuits and central pattern generators. These basic forms of activity may have been sufficient for the earliest vertebrates that swam in the ocean: jawless scavengers like the hagfish. However, some of these ver- tebrates gradually developed a new and more active lifestyle. They evolved jaws for capturing and consuming mobile prey, along with improved sensory organs for locating the prey. With this new lifestyle came a need for enhanced motor control.
Shaping the raw motor activity of reflexes and central pattern generators into smooth, efficient movements is a
02-Eagleman_Chap02.indd 54 02/11/15 3:16 pm
The Cerebellum 55
# 158305 Cust: OUP Au: Eagleman Pg. No. 55 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.18 The cranial nerves and their major functions.
(VIII) Vestibulocochlear Carries signals for the senses of hearing and balance
(III) Oculomotor Controls eye movement and pupillary constriction
(IV) Trochlear Controls eye movement
(VI) Abducens Controls eye movement
III
VI
IV (V) Trigeminal Controls the muscles of mastication (chewing); involved in the sensation of touch and pain by the face and mouth
(VII) Facial and intermediate Carries signals for the sense of taste (anterior 2/3 of tongue); controls the muscles of facial expression; involved in the secretion of tears and saliva
(IX) Glossopharyngeal Carries signals for the sense of taste (posterior 1/3 of tongue); mediates the swallowing reflex
(I) Olfactory Carries signals for the sense of smell from the nasal passage to the brain
(XI) Spinal Accessory Controls some muscles for movements of the head, neck, and shoulders
(XII) Hypoglossal Controls muscle of the tongue
Sensory fibers Motor fibers
(II) Optic Carries visual signals from the retina to the thalamus
(X) Vagus A major input and output pathway for parasympathetic nervous system; senses aortic blood pressure; slows heart rate; stimulates digestive organs
V
VII
VIII
IX
X
XI
XII
I II
Intermediate nerve
02-Eagleman_Chap02.indd 55 02/11/15 3:16 pm
56 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 56 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
complicated process. Movements take place in an environ- ment and hence require steering: toward targets and away from obstacles. Proper steering cannot be blind—it needs sensory input for guidance. W hen a movement veers off its intended path, it needs correction. If the target tries to flee, the movement needs adjustment. A ll of these functions re- quire adding on a new kind of circuitry, and lots of it. This circuitry lies in the cerebellum.
Circuitry of the “Little Brain” The cerebellum, or “little brain,” contains an enormous number of neurons. The human cerebellum contains more neurons than both hemispheres of our much larger cere- brum combined, and in many species the cerebellum dwarfs all other brain structures. A ll these neurons are densely packed into leaflike folia, which are themselves packed into larger folds or lobules, which in turn are packed into larger lobes (FIGURE 2.19). This arrangement fits as many neurons into as small a space as possible, allowing them to communi- cate more efficiently. The cell bodies of these neurons are found in a wrinkly sheet covering the outer surface or cere- bellar cortex (not to be confused with the cerebral cortex, which we’ll discuss later in the chapter).
The microcircuitry of the cerebellar cortex has a remark- ably consistent wiring pattern across the entire sheet, in all species of vertebrates (FIGURE 2.20). Inputs come from
FIGURE 2.19 The gray matter of the cerebellum is densely packed into leaflike structures called folia, lobules, and lobes.
Cerebellar cortex
Folia
Lobe
Lobule Lobule
FIGURE 2.20 The microcircuitry of the cerebellum.
Molecular layer
Granule cell layer
White matter
Purkinje cell layer
Purkinje cell
Mossy fiberClimbing
fiber
Purkinje cell axons
Granule cell
Granule cell axons
Excitatory neurons
Inhibitory neurons
Inputs (mossy and climbing fibers)
Outputs (Purkinje cell axons)
dedicated nuclei in the brainstem and connect to small excit- atory and inhibitory interneurons in the lower part of the sheet, called the granule cell layer. These interneurons send their output signals to the upper part of the sheet, called the molecular layer. Sandwiched between these two layers lie the giant output neurons, named Purkinje cells after the Czech anatomist who discovered them. Purkinje cells have a beautiful, intricate branchwork of input connections that gather information from the molecular layer above them. They integrate this information and send their output back to specialized output nuclei in the brainstem, which pass the information back to the spinal cord and ahead to the cerebral cortex and the rest of the brain.
Functions of the Little Brain In the brain, function follows circuitry, so the remarkable consistency of cerebellar circuitry across species suggests a common and important function. W hat could it be? Damage
02-Eagleman_Chap02.indd 56 02/11/15 3:17 pm
The Diencephalon: Hypothalamus and Thalamus 57
# 158305 Cust: OUP Au: Eagleman Pg. No. 57 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
to the cerebellum interferes with the smooth, efficient move- ments of body parts to their targets in the surrounding envi- ronment. Movements become jerky and clumsy. They overshoot or undershoot their targets, occur too early or too late, or are too strong or too weak. Maintaining upright bal- ance becomes difficult or impossible. Learning new motor responses also becomes more difficult.
The cerebellum clearly has an important role in coordi- nating movements and matching them to their environment, although neuroscientists debate exactly what this role might be. One possibilit y is that the cerebellum calculates a forward model of upcoming movements (Seidler, Noll, & Thiers, 2004). In other words, it makes predictions about the expected sensory outcomes of motor actions and uses these predictions to refine outgoing motor commands. Building a forward model turns out to be incredibly useful for planning fast, precise movements. For example, when you are run- ning, there is not enough time for your brain to send out a motor command to your foot, then wait for the sensory feed- back to return to see what happened, and then send out an- other motor command to adjust the positioning. You would lose your balance and fall over immediately if the system were this slow. Even when you are standing still, the time delay between your vestibular organ sensing how far your legs are from vertical and the motor command arriving at your leg muscles to compensate is too long to maintain bal- ance without a forward model. Likewise, catching a ball or shaking a friend’s hand is impossible unless you aim your movements at where the target is going to be, not at where it was a moment ago. By planning ahead, the cerebellum allows a whole new level of sophistication in motor control.
The cerebellum was traditionally considered solely a motor structure. However, more recent research has also found evidence that the cerebellum plays an important role in nonmotor functions such as language, memory, attention, and even emotional regulation (Schmahmann, 2010; Stood- ley, Valera, & Schmahmann, 2012; Strick, Dum, & Fiez, 2009). This is an excellent example of how our understand- ing of brain function is always evolving.
The Diencephalon: Hypothalamus and Thalamus As we enter the forebrain, the nervous system takes on quite a different structure than we have seen until this point. Here we find few direct inputs or outputs to the body or the out- side world. Instead, the neurons of the forebrain can mostly be considered interneurons, whose vast and elaborate cir- cuitry is added on to the simpler circuits we have reviewed so far. Recall that the brain adds complexity to behavior by in- serting layers of processing between sensory input and
motor output. The additional neural circuits of the forebrain thus tend to have quite complex roles—roles that can be quite flexible over time. We’ll begin our look at these circuits with the two main structures of the diencephalon, which lies just forward of the midbrain circuits that we discussed in the section on the brainstem. These structures are the hypo- thalamus and the thalamus.
hypothalamus: A Keystone Structure in homeostasis A ll living organisms have survival needs. The body remains alive only within a narrow range of physical parameters. With insufficient energy supplies, or too little water, or too much water, or too high a temperature, or too cold a tem- perature, all the delicately balanced biochemical processes that sustain life will grind to a halt. Brains that allow the body’s internal parameters to wander too far from the ideal range do not allow the body to survive to reproduce. A fter millions of generations of experience, the brains of today’s living creatures have gotten very, very good at maintaining homeostasis (the process of keeping the body’s internal pa- rameters in balance). The neurons that drive homeostasis can be found in the hypothalamus.
Hypothalamic neurons are responsible for the homeo- static control signals we sometimes call “basic drives.” Basic drives include hunger, thirst, sexual arousal, temperature reg- ulation, and sleep. These drives serve to maintain the body’s balance of energy intake against energy consumption, water intake against dehydration, temperature regulation against overheating and overcooling, and so on. Neurons of the hypo- thalamus cluster into distinct groups: the hypothalamic nuclei (FIGURE 2.21). Each hypothalamic nucleus has a distinct func- tion, and many relate to a specific drive. For example, one nu- cleus coordinates feeding; another regulates satiety; others regulate heat-generating behavior, heat-shedding behavior, and mating behavior. Since many drives wax and wane accord- ing to the time of day or night, one hypothalamic nucleus acts as a circadian clock to stimulate or inhibit the other nuclei.
To maintain homeostasis, the neurons of the hypothala- mus need input about the internal state of the body. They obtain this information from many sources: visceral inputs via the spinal cord, hormonal inputs from other body organs, even direct measurements of the chemistry of the bloodstream. They integrate the information from all of these sources and compare the results against ideal homeostatic set points. The set points themselves can be changed when necessary. For ex- ample, internal body temperature can be increased to fight in- fection. Hunger can increase, leading to increased food consumption and storage of energy for the winter months.
W hen the internal environment deviates too far from the set point, the hypothalamus coordinates the necessary com- pensatory mechanisms. These compensatory mechanisms fall into three categories: autonomic responses, endocrine
02-Eagleman_Chap02.indd 57 02/11/15 3:17 pm
58 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 58 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.21 Nuclei of the hypothalamus, with their major functions.
Mamillary body
Dorsomedial nucleus
Heat dissipation Heat conservation
Medial preoptic nucleus
Anterior nucleus
Suprachiasmatic nucleus
Arcuate nucleus
Ventromedial nucleus
Posterior nucleus
Lateral hypothalamic area
Paraventricular nucleus
Circadian rhythms Appetite and thirst Satiety Hormonal regulation
responses, and behavioral responses. Hypothalamic neu- rons send outputs via the thalamus to the cerebral cortex, which has the computational power to elaborate basic drives into goals and plans of action. (The next section of the chap- ter describes the cerebral cortex in greater depth.) The hypo- thalamus also provides extensive output to the autonomic control centers of the brainstem and spinal cord.
Furthermore, the hypothalamus is considered the master control gland of the body’s hormone-secreting systems (which are known collectively as the neuroendocrine system) (FIGURE 2.22). It sends control signals down a thin extension to the pituitary gland, which in turn releases many kinds of hormonal signals into other parts of the body. One example of a pituitary hormone is growth hormone, which regulates tissue growth throughout the body. Thyroid-stimulating hormone directs the thyroid in controlling the body’s over- all metabolic rate. Prolactin regulates lactation. Ox y tocin facilitates maternal bonding, lactation, and social bonding. Antidiuretic hormone directs the kidneys to retain rather than excrete water. These kinds of hormonal responses are essential components of homeostasis. We will learn more about the hypothalamus and its functions in Chapter 13.
How might homeostasis work in practice? Imagine that an organism has not consumed water in several days and becomes dehydrated. The hy pothalamus may sense that the blood pressure is dropping and that the osmolarity (a measure of the total concentration of sodium, potassium, and other dissolved chemical constitutents) of its bodily
fluids is increasing to unacceptable levels. It can compen- sate by sending signals to the autonomic control centers of the brainstem and spinal cord to increase the heart rate and constrict the blood vessels; this will help to maintain blood pressure. It can also stimulate the pituitar y gland to produce antidiuretic hormone, which signals the k idneys to stop excreting water. These k inds of physiological com- pensator y mechanisms will buy the organism some time. However, the autonomic and endocrine responses are not enough to ensure sur vival without the accompanying be- havioral component. The organism needs to find some water, quick ly!
By itself, the hypothalamus is not equipped to actually find or consume any water. Foraging for water sources is an extremely demanding task. It requires multisensory input, a repository of past memories, a weighing of possible plans of action, and complex motor behavior. The hypothalamus lacks the necessary circuitry to perform any of these tasks. The cerebral cortex, however, is readily capable of perform- ing all the necessary steps for water finding, just as soon as it receives a motivational signal to begin. So the hypothalamus generates a motivational alarm signal and then passes the signal to the cerebral cortex to decide what to do about it. The cerebral cortex is a large and complicated place, however. Getting the signal to the appropriate areas of the cerebral
02-Eagleman_Chap02.indd 58 02/11/15 3:17 pm
The Diencephalon: Hypothalamus and Thalamus 59
# 158305 Cust: OUP Au: Eagleman Pg. No. 59 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
regions to work together (X. J. Wang & R inzel, 1993). This role in synchronization may be important for attention, awareness, and the conscious state itself (Bartlett & Wang, 2011; Guldenmund et al., 2013; Schmidt, 2003; Zhang & Bertram, 2002).
The thalamus and the cerebral cortex are tightly inter- connected. Each thalamic nucleus serves a different section of the cerebral cortex. For example, the lateral geniculate nucleus relays information from the light-sensitive neurons (or photoreceptors) of the retina to the neurons of the pri- mary visual cortex (V1), where the first stages of visual in- formation processing in the cortex take place. Another thalamic nucleus, the pulvinar nucleus, conveys the signals from the superior colliculus (which, as you may recall, is in- volved in moving the eyes to fixate on visual stimuli) to areas of the cerebral cortex that perform similar functions. Other thalamic nuclei convey auditory information from the brain- stem to auditory areas of the cerebral cortex, tactile informa- tion to the somatosensory cortex, and visceral information (including the sense of taste) to visceral sensory areas of the cerebral cortex.
Still other thalamic nuclei serve motor control regions of the cerebral cortex. Some signals from the cerebellum do not pass to the spinal cord, but instead find their way to the cere- bral cortex via relay stations in the thalamus (FIGURE 2.24). This allows the cerebral cortex to use the future-predicting “forward models” of the cerebellum for its own motor func- tions. Other motor signals pass to the cerebral cortex from the substantia nigra in the brainstem or from the basal gan- glia (motor control structures anterior to the thalamus, about which you will learn in the next section). Thalamic nuclei provide the motor cortex with access to all of these diverse sources of motor control information.
Other thalamic nuclei are dedicated to serving associa- tion areas of the cerebral cortex. The association areas of the
FIGURE 2.22 Hormone-regulating functions of the hypothalamus. Nuclei in the hypothalamus send signals down to the pituitary gland to control the secretion of a variety of key hormones for controlling the functions of the internal organs of the body.
Kidney
Adrenal cortex
Thyroid
OvariesTestis
Kidneys retain water
ACTH
ADH
Thyroid stimulating hormone
Growth hormone
Oxytocin
Prolactin
FSH and LH
Bone and tissue
Uterine contractions
Lactation and Maternal bonding
Anterior pituitary
Posterior pituitary
cortex is a job all in itself. The neurons that take on this task lie in the other major diencephalic structure: the thalamus. Let’s now turn to this structure and see how it works.
Thalamus The thalamus plays a central role in brain function by acting as a relay station to the cerebral cortex, conveying incoming sensory information to the appropriate cortical areas. It also relays motor signals to the cerebral cortex from other motor control structures like the cerebellum and basal ganglia (which we’ll look at in the next section). In addition, the thal- amus acts as a relay station between distant areas of the cere- bral cortex itself, communicating information from one area to another. Thalamic neurons cluster into a large number of separate thalamic nuclei, which serve different regions and therefore play different roles (FIGURE 2.23). Further, the tha- lamic nuclei may play an important role in sy nchronizing neural activ it y bet ween distant regions, enabling these
FIGURE 2.23 The thalamus is divided into a number of nuclei. Relay nuclei transmit information to and from specific regions of the cerebral cortex. Intralaminar nuclei connect diffusely to large areas of cortex. The reticular nucleus wraps around the other nuclei and regulates their activity.
Relay nuclei
Intralaminar nuclei
Internal medullary lamina
Reticular nucleus
02-Eagleman_Chap02.indd 59 02/11/15 3:17 pm
60 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 60 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
consciousness in general (Edelstyn, Mayes, & Ellis, 2014). Even a small injury to the intralaminar nuclei can produce a profound reduction in the level of consciousness.
The reticular nucleus is one final and very important structure of the thalamus. It consists of a thin sheet of neu- rons that wraps around the entire surface of the thalamus. Unlike all other thalamic nuclei, it has no connections to the cerebral cortex and few inputs from any other outside brain structure. Instead, almost all of the input to the reticular nu- cleus originates within the thalamus itself. Its neurons, all of which are inhibitory, connect only to other nuclei of the thal- amus. Each neuron not only inhibits some of the neurons of a thalamic nucleus, but also inhibits its own neighbors. The effect of this is like changing a babble of uncoordinated neural chatter into a respectful conversation, where other neurons grow silent when one of their neighbors is speaking. The re- ticular formation may help to organize the communication activity of the thalamic nuclei themselves. The brainstem re- ticular formation stimulates the activity of the reticular nu- cleus, so that its activity waxes and wanes over the course of the day and night. As it does so, the level of alertness and the clarity of consciousness increase and decrease. Attention, awareness, and consciousness all critically depend on the in- formation-conveying capacity of the thalamus.
cerebral cortex are neither purely sensory nor purely motor areas. These areas integrate sensory and motor functions and are important for more complex forms of sensory pro- cessing and motor planning.
Additional thalamic nuclei convey information about motivations and drives to the cerebral cortex. Earlier in this section, we saw how the basic homeostatic drives of the hy- pothalamus must be unpacked into priorities, goals, and action plans so that the organism’s behavior meets its essen- tial needs. Specialized nuclei of the thalamus relay informa- tion from the hypothalamus and other key motivational structures to the cerebral cortex for behavioral planning.
Other types of thalamic nuclei, called the intralaminar nuclei, provide a more diffuse input to large swaths of the cerebral cortex as a whole. Since they connect to so many areas with such diverse functions, these nuclei are probably not involved in any single type of sensory or motor function (Benarroch, 2008; Van der Werf, Witter, & Groenewegen, 2002). Much of their input comes from brainstem structures involved in alerting and arousal, such as the reticular forma- tion that we discussed in the section on the midbrain, which plays a central role in switching the brain from the conscious to the unconscious state. In addition, these small thalamic nuclei seem to be important in maintaining alertness and
FIGURE 2.24 The relay nuclei of the thalamus are tightly connected to the cerebral cortex. Each relay nucleus serves a specific region of cortex and helps it to perform a specific function.
Cingulate gyrus
Lateral geniculate nucleus Somatosensory
cortex
Motor and premotor cortex
Medial geniculate nucleus
Auditory cortex
Visual cortex
Prefrontal cortex Parietal cortex
Pulvinar VPM
MD
VL
VA
LD
LP
Ant.
VPL
02-Eagleman_Chap02.indd 60 02/11/15 3:17 pm
The Telencephalon: Cerebral Cortex and Basal Ganglia 61
# 158305 Cust: OUP Au: Eagleman Pg. No. 61 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
CASE STUDY: Waking the Brain Around the turn of the millennium, a young man was assaulted and beaten unconscious while being robbed. He suffered such severe brain injury in the attack that he re- mained in a diminished state of con- sciousness for more than two years. He made no signs of verbal commu- nication, rarely responded to com- mands, and had severely spastic limbs. At that point, with little hope for further recovery, he was trans- ferred from a hospital to a long- term care facility. More than six years after the injury, his condition had not improved much. However, a new series of tests gave some cause for optimism. Brain imaging showed that the network of regions for lan- guage function was still intact; other networks might have been similarly spared. Overall, his brain showed much less metabolic activity than a normal, awake brain. Yet perhaps much of his brain was merely dor- mant rather than destroyed.
In brain injury, the intralaminar nuclei of the thalamus are sometimes
damaged to the point of causing un- consciousness although the rest of the brain’s circuitry is not as pro- foundly affected. If the patient could only be “awakened,” many brain func- tions might return. In an attempt to awaken the patient, a team of neurol- ogists and neurosurgeons led by Dr. Nicholas Schiff implanted the intrala- minar nuclei of the patient’s thalamus with the electrodes of a deep brain stimulator (Schiff et al., 2007; Schiff & Posner, 2007).
Deep brain stimulation involves the surgical implantation of elec- trodes into the brain. These elec- trodes send electrical impulses to specified parts of the brain. A pace- maker-like device that is placed under the patient’s skin in his or her upper chest controls the amount of brain stimulation. A wire that trav- els under the patient’s skin con- nects this pacemaker-like device to the implanted electrodes in the brain.
The results of the stimulation were striking. After being virtually
unresponsive for nearly seven years, the patient began opening his eyes and turning his head to the sound of a voice. He became able to swallow food placed on his tongue and even began to speak some simple sen- tences. His condition continued to improve over the following months. He began communicating reliably with his family and the medical staff, watching movies, and even laughing at appropriate times. Since then, other patients in an altered state of consciousness have also undergone the procedure. Although deep brain stimulation is not always so suc- cessful, some of these patients have shown similar improvements in their level of consciousness (Giacino, Fins, Machado, & Schiff, 2012). Al- though resuming a completely inde- pendent life is rarely possible after such severe injuries, in some cases deep brain stimulation can return these patients to their families and to the world of waking life.
The Telencephalon: Cerebral Cortex and Basal Ganglia
Cerebral Cortex The cerebral cortex is the largest part of the human brain and the most quintessential. W hen we picture a human brain in our minds, we tend to imagine a pinkish, wrinkly organ look- ing vaguely like an oversized walnut out of its shell. This con- voluted structure is actually not the entire brain, but simply the outer covering: the cerebral cortex (cortex is Latin for bark,
as in tree bark). Present in all mammals but dramatically ex- panded in humans, the cerebral cortex is critical for all of the most elaborate forms of human cognition: speaking a sen- tence, reading the words on a page, planning goals for the future, turning those goals into actions, recognizing and using tools, imagining the future and the past, thinking about what other people are thinking, and being aware of our own selves. Since the neuroscience of cognition is the subject of this entire book, we’ll have lots of time to consider each of these advanced functions in detail in the chapters ahead. For now, we’ll just familiarize ourselves with the different parts of the cerebral cortex and their general functions (FIGURE 2.25).
Early neuroanatomists could only speculate as to the functions of the different parts of the cerebral cortex. They looked at the brain’s overall appearance, divided up the structures they saw in front of them, and assigned names to each of them. Of course, the brain was under no obligation to
02-Eagleman_Chap02.indd 61 02/11/15 3:17 pm
62 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 62 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.25 The major lobes, landmarks, gyri, and sulci of the cerebral cortex. (a) The major lobes and landmarks; (b) lateral view of the left cerebral hemisphere; (c) medial view of the right cerebral hemisphere.
Frontal lobe
Parietal lobe
Occipital lobe
Temporal lobe
Limbic lobe
(a)
(c)
(b)
Central sulcus
lateral sulcus
Parieto- Occipital sulcus
Cingulate sulcus
Superior
Inferior
PosteriorAnterior
Superior frontal sulcus Postcentral sulcus
Postcentral gyrus
Paracentral lobule
Precuneus
Cuneus
Occipital gyri
Calcarine sulcus
Intraparietal sulcus
Superior parietal lobe
Inferior parietal lobe
Angular gyrus
Supramarginal gyrus
Precentral gyrus
Superior frontal gyrus
Cingulate gyrus
Dorsal anterior cingulate cortex
Posterior cingulate cortex
Mid-cingulate cortex
Retresplenial cingulate cortex
Subgenual cingulate gyrus
Pregenual cingulate gyrus
Gyrus rectus
Superior frontal gyrus
Middle frontal gyrus
Precentral sulcus
Inferior fontal gyrus
Orbitofontal gyri
Superior temporal gyrus
Middle temporal gyrus
Inferior temporal sulcus
Inferior temporal gyrus
Superior temporal sulcus
Temporal pole
Inferior frontal sulcus
Corp us callosum
02-Eagleman_Chap02.indd 62 02/11/15 3:17 pm
The Telencephalon: Cerebral Cortex and Basal Ganglia 63
# 158305 Cust: OUP Au: Eagleman Pg. No. 63 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Each hemisphere is composed of four smaller lobes: the frontal lobe, temporal lobe, parietal lobe, and occipital lobe. A large lateral sulcus runs along the side of each hemi- sphere of the cerebral cortex; below it lies the temporal lobe (under what we call the “temples” of the head). Above it, a vertical central sulcus lies between the frontal lobe and the more posterior parietal lobe. Behind the parietal and tempo- ral lobes, at the back of the cerebral cortex, lies the occipital lobe. Each of these lobes has a variety of functions, which we’ll consider in a moment. In addition, each lobe has fur- ther anatomical and functional subdivisions.
Broadly speaking, the dividing line between the “front” and the “ back ” of the cerebral cortex is the central sulcus. In rough terms, everything in front of the central sulcus does various forms of motor planning and action, whereas every- thing behind and below the central sulcus does various forms of sensory processing. Just in front of the central sulcus lies the precentral gyrus, which is home to the pri- mary motor cortex: a long strip of areas that controls move- ments of individual body parts. In front of this gyrus are areas involved in planning movements. In front of those areas is the prefrontal cortex, which assembles more elabo- rate sequences of movement and behavior and is a major player in cognition and goal planning. The prefrontal cortex has a superior, middle, and inferior frontal gyrus on its lat- eral side. The medial prefrontal cortex lies along the medial wall of the frontal lobe. The underside of the prefrontal cortex, above the orbits of the eyes, is called the orbitofron- tal cortex. It plays an important role in setting priorities and determining how valuable an action or a resource might be, given current needs. The olfactory cortex, which is impor- tant in the sense of smell, also lies in this area.
Just behind the central sulcus lies the postcentral gyrus, which is home to the primary somatosensory cortex (S1): another strip of areas that handles sensory input from the skin, muscles, and joints of individual body parts. Behind it, the rest of the parietal lobe is divided into the inferior and superior parietal lobules. The dividing line between these lobules is the intraparietal sulcus, in which a large area of ce- rebral cortex is hidden. Superior parts of the parietal lobe play a key role in locating objects in space, like a more elabo- rate version of the midbrain’s superior and inferior colliculi. This is useful for planning where to make movements in space. Inferior parts of the parietal lobe play a role in orga- nizing stimuli according to their form rather than their loca- tion. This is useful for planning what kind of movements to make. The final part of the parietal lobe lies on the medial wall and is called the precuneus. It is one of the most active regions of the brain, even when we are at rest. It is active when we are imagining scenes and when we are navigating: thinking of destinations and finding directions to them.
Behind the parietal lobes lie the gyri of the occipital lobe. The occipital lobe is devoted to processing visual input and contains many different subregions for mapping out the vari- ous features of visual stimuli: position, orientation, shape, color, motion, and so on. The primary visual cortex lies on
organize its functions in the same way! The result is that we have a long-standing system of anatomical names that does not always map neatly to the functional organization of the cerebral cortex itself. Some structures sprawl across many kinds of functional circuits, whereas others do not map well to any particular function.
The cerebral cortex consists of a layered, outer sheet of gray matter surrounding an inner white matter. As in other parts of the nervous system, the gray matter is composed mostly of the cell bodies of neurons and their local connec- tions, and the white matter is composed of long-distance connection fibers linking neurons that are distant from one another. On the surface, the rounded convolutions of the ce- rebral cortex are called gyri (singular, gyrus), and the grooves between gyri are called sulci (singular, sulcus). As with the cerebellum, this crumpled pattern of gyri and sulci allows the brain to fit a large sheet of cerebral cortex into a small space while minimizing the distance between any two neurons (FIGURE 2.26).
A large midsagittal sulcus (also known as the longitudinal fissure) divides the cerebral cortex into left and right hemi- spheres, which have a lateral and a medial wall. On the medial wall, you can see the large bridge of white matter connections between the two hemispheres: the corpus callosum. The corpus callosum allows the left and right hemispheres to communicate with one another.
FIGURE 2.26 Major structures of the ventral (underside) surface of the cerebral cortex, with brainstem removed.
Temporal pole
Inferior temporal gyrus
Collateral sulcus
Fusiform gyrus (occipitotemporal gyrus)
Occipitotemporal sulcus
Parahippocampal gyrus
Anterior
Posterior
LeftRight
02-Eagleman_Chap02.indd 63 02/11/15 3:17 pm
64 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 64 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
the medial wall of the occipital lobe, mostly tucked away inside the deep calcarine sulcus.
On the medial wall lies the visceral motor cortex of the cingulate gyrus, which wraps like a belt (Latin, “cingu- lum”) around the hemisphere-spanning bridge of the corpus callosum. The cingulate gyrus is involved in many different functions, as we’ll see in a moment.
Below the lateral sulcus lies the temporal lobe, with a su- perior, middle, and inferior temporal gyrus on its lateral side. The superior temporal cortex handles auditory infor- mation, with the primary auditory cortex (A1) tucked just inside the posterior part of the lateral sulcus. The underside of the temporal lobe has two more gyri: the fusiform gyrus and parahippocampal gyrus. These areas are associated with a ventral visual pathway that handles the identification, cate- gorization, and evaluation of visual inputs: faces, houses, cars, animals, and other objects in the surroundings (Mahon et al., 2007; Q. Wang, Sporns, & Burkhalter, 2012).
Hidden away within the depths of the lateral sulcus lies a large area of cerebral cortex known as the insula (Latin, “island”). The insula is the visceral sensory part of the cere- bral cortex. It represents the state of the internal organs and registers internal bodily states like pain, fatigue, hunger, sexual arousal, and so on (FIGURE 2.27). Taste receptors send input to the primary gustatory cortex, which lies in this area.
Basal Ganglia The basal ganglia are a set of closely interconnected gray matter structures beneath the white matter of the cerebral
FIGURE 2.27 The insula region of the cortex is hidden inside the lateral sulcus. It represents the internal or visceral sensations of the body.
Insula
Lateral sulcus (retracted)
RESEARCH METHODS: Cytoarchitecture of the Cortex Biologists have made tremendous progress in learning about the cir- cuitry of the nervous system. The word cytoarchitecture refers to the microscopic structure of the circuits of neurons within a part of the ner- vous system (FIGURE 2.28). In general terms, the cytoarchitecture of the cerebral cortex has many common features both across areas of the cortical sheet and across mamma- lian species. Most parts of the cere- bral cortex have six layers, each with characteristic connections to other
areas of the cerebral cortex, thala- mus, cerebellum, and other parts of the nervous system. The relative thickness and appearance of each layer are slightly different in differ- ent cortical areas.
The subtle differences in the cy- toarchitecture of different cortical regions can help us in mapping the functional anatomy of the brain. In a classic work in 1909, the German neuroanatomist Korbinian Brod- mann created a map dividing the ce- rebral cortex into 52 distinct regions
for nonhuman primates (and 43 dis- tinct regions for humans), each with a different characteristic cytoarchi- tecture (Zilles & Amunts, 2010). He also created similar maps in other species. Although he did not assign different roles to these different re- gions, later investigators found that the borders often mapped fairly well to specific functions: Brodmann’s area 17 corresponded to primary visual cortex, area 4 corresponded to primary motor cortex, and so on (Zilles & Amunts, 2010).
cortex (FIGURE 2.29). They play an important role in initiating and maintaining activity in the cerebral cortex, particularly in the motor control areas of the frontal lobes, which must often be driven by an organism’s internal goals and needs. For example, the basal ganglia are involved in a diverse set of functions: limb movements, eye movements, planning and goal setting, motivation, and reward. We will consider the basal ganglia here in a brief overview.
The outermost structure is called the striatum. It con- sists of a comet-shape structure called the caudate nucleus
02-Eagleman_Chap02.indd 64 02/11/15 3:17 pm
The Telencephalon: Cerebral Cortex and Basal Ganglia 65
# 158305 Cust: OUP Au: Eagleman Pg. No. 65 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Neuroanatomists have contin- ued to use differences in circuitry to distinguish functional areas in the cerebral cortex. Some maps divide the cerebral cortex into a different zone for each nucleus of the thalamus. Others look at pat- terns of long- distance connections to other brain regions, at the order in which the areas mature during development, or at the different kinds of receptors and proteins in each region (using immunohisto- chemical techniques, which employ antibodies as tracers of receptor or protein distribution). The most contemporary methods use an au- tomated process in which the bor- ders between areas are detected by computer algorithms reading mi- croscopic slices of brain tissue. These slices are then reassembled to build maps of the brain. The maps can be compared across many indi- viduals to understand not only the similarities but also the differences in brain architecture. Whatever method is used, cytoarchitectonic maps are based on the principle that the circuitry tells us something important about the function. By understanding these different types of circuits, neuroanatomists hope to unravel the puzzling fabric of the brain’s many functions.
FIGURE 2.28 The cortex can be divided into subregions based on cytoarchitecture: the microscopic appearance of the neural circuitry. Several cytoarchitectonic maps of the brain are available. The most widely used is the Brodmann atlas, which divides the cortex into about 50 numbered regions.
I II
III
IV
V
VI
I II
III
IV
V
VI
I
II
III
IV
V
VI
Prefrontal association cortex (area 46)
Primary motor cortex (area 4)
Primary visual cortex (area 17)
1
2 5
7
39 40
19
3722 21
20
38
42 41
43
4 6
8
9
44
45
47
10
11
46
18
17
3
and a round structure called the putamen, sitting within the “C” of the caudate. These two nuclei actually begin as a single structure, but as the brain develops, they become separated by the internal capsule: a massive tract of white matter heading from the cerebral cortex down to the spinal cord, brainstem, cerebellum, and thalamus. By the end of the fetal brain development process, the caudate nucleus and the putamen are connected by only a few thin stripes of gray matter: “striae,” hence the collective term “striatum.” The ventral striatum contains a structure known as the nu- cleus accumbens—an important player in reward and ad- diction, as we’ll see in Chapter 14. Underneath the putamen lies another ovoid structure called the globus pallidus (Latin: “pale globe”), a critical area for regulating voluntary movement.
Other nearby structures work closely with the basal gan- glia, even if they are not always considered under that
umbrella term. Continuing inward from the globus pallidus, under the thalamus, we find the subthalamic nucleus. Below the subthalamic nucleus, we find the midbrain’s sub- stantia nigra. These areas are well connected to the basal ganglia, and they participate in the same functions.
The neurons of the basal ganglia are densely intercon- nected with the cerebral cortex, especially with the frontal cortex. Cortical neurons send connections down to the stri- atum, which in turn sends connections further inward to the internal and external globus pallidus, sometimes indirectly through the subthalamic nucleus. From here, the circuit continues on to the motor nuclei of the thalamus, which con- nect back to the original site of the cerebral cortex, forming a complete loop. Different loops or channels serve different regions of the cerebral cortex, which means that they have different functions, as mentioned earlier. We’ll take a closer look at these functions in later chapters.
02-Eagleman_Chap02.indd 65 02/11/15 3:17 pm
66 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 66 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.29 Major structures of the basal ganglia. (a) The basal ganglia consist of several gray matter structures beneath the gray and white matter of the cortex. (b) Loop circuits from the cortex through these structures allow the brain to initiate, maintain, and terminate various classes of thought and behavior.
Thalamus
From frontal cortex
To frontal cortex
Globus pallidus Internal globus pallidus
Subthalamic nucleus
External globus pallidusPutamen
Caudate nucleus
Subthalamic nucleus
Substantia nigra
(a) (b)
Simple movements
Primary motor loop
Complex movements
Eye movements
Cognition
Reward, Evaluation
Orbitofrontal loop
Dorsolateral loop
Oculomotor loop
Premotor loop
Uniting the Inside and Outside Worlds
The Limbic System As you have seen throughout this chapter, the nervous system contains distinct input and output pathways for deal- ing with the external environment and the internal environ- ment. The brain takes input from the external environment via visual inputs from the eyes, auditory inputs from the ears, touch and position inputs from mechanical receptors in the skin and joints, and so on. It also takes input from the inter- nal environment of the heart, lungs, blood and blood vessels, and other visceral tissues via a diverse set of pain, tempera- ture, itch, chemical, and pressure sensors known as intero- ceptors (Craig, 2002). Likewise, the brain’s output pathways include both motor neurons for controlling the muscles, joints, and skeleton to allow the body to take actions within the external environment and the neurons of the autonomic nervous system (both sympathetic and parasympathetic di- visions) for controlling the internal environment of the heart, lungs, blood and blood vessels, and other visceral tissues.
The internal environment (“inner world”) must, how- ever, interface with the external environment (“outer world”) at some point in the nervous system. At every level of the central nervous system, we can find regions where the sen- sory inputs from both internal and external environments converge to help guide control of the internal environment (Nauta, 1979; Nieuwenhuys, 2008). These areas are some- times considered as forming their own system, central to motivation and emotion. They are known as the limbic system, from the Latin limbus or “ border” (FIGURE 2.30).
The very concept of the limbic system is an old one, dating back to the middle of the 19th century (McLachlan, 2009). Over the years, most neuroscientists have agreed that the general concept of a brain system for emotion and motivation is a useful one. However, they have had difficulty agreeing on the details. Rarely do two books on neuroanat- omy agree on exactly which brain structures are and are not part of the limbic system. Here we will take a look at some of the most agreed-on structures whose circuits bridge the gap between the brain’s internal-environment sensory inputs and motor outputs. These structures have continuously
02-Eagleman_Chap02.indd 66 02/11/15 3:17 pm
Uniting the Inside and Outside Worlds 67
# 158305 Cust: OUP Au: Eagleman Pg. No. 67 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 2.30 The limbic system is a set of regions involved in generating the hormonal, autonomic, and motivational aspects of emotional states. Core limbic structures include the hypothalamus, amygdala, and hippocampus, as well as some nuclei of the midbrain and brainstem. They also work closely with certain circuits through the basal ganglia and cortex.
Other cortical regionsOther cortical regionsOther cortical regions
Limbic cortex
Ventrolateral medulla
Brainstem Amygdala
Thalamus
Basal ganglia Basal ganglia
Hypothalamus
Olfactory bulb
Hippocampus Nucleus solitarius
Spinal cord, cranial nerves and neuro- hormonal pathways
Parabrachial nucleus
Periaqueductal gray
Septal area and basal forebrain
Septal area and basal forebrainb iib ii
pppppeptal area and basad b Septal area and basal forebrain
proven themselves critical in the regulation of motivation and emotion.
The hy pothalamus plays a key role in homeostasis and in motivation, as we saw earlier (for example, the motiva- tion to go find water when thirsty). It is considered a core structure of the limbic system. Parts of the thalamus convey hy pothalamic signals to the cerebral cortex; hence, these nuclei are sometimes called limbic nuclei of the thalamus. Certain parts of the midbrain, such as the periaqueductal gray, also link together visceral and somatic functions to produce simple k inds of motivated and emotional behavior. A lso in the midbrain, substantia nigra neurons contain do- pamine, a neurotransmitter that is central to motivation and reward. Neurons of the raphe nuclei contain serotonin,
a neurotransmitter that is also important for emotion and the regulation of internal states.
In our discussion of the hypothalamus, we saw how in- ternal bodily states are the basis of motivations and drives: they reflect internal needs, but lead to external-world behav- ior. Emotions, too, are built on a foundation of internal bodily states. Imagine a time when you felt fear, anger, love, friendship, pride, embarrassment, shame, disgust, content- ment, or joy: all of these emotions are underpinned by strong visceral sensations. Emotions can be used to assign value to otherwise neutral sensory stimuli from the external environ- ment: a face is suddenly recognized as a friendly face or a house is recognized as one’s own home.
In addition, two important limbic structures lie in the medial temporal lobes. The first of these is the amygdala (Greek for “almond,” whose shape it vaguely resembles). The amygdala is similar to the hypothalamus in its outputs: it, too, can directly drive the internal states of the body through auto- nomic mechanisms and hormonal signals. It also sends outputs to the cerebral cortex to drive motivated behaviors, prioritiza- tion, goal setting, and action planning. However, its inputs are different from those of the hypothalamus. Rather than draw- ing on the inside world, the amygdala obtains input directly from the external-world senses of vision, hearing, and smell.
The amygdala is a necessary complement to the hypothal- amus. After all, an organism whose only drives came from its own internal organs would not react to the external threat of a predator or a rival. Nor would it appreciate the value of an ex- ternal opportunity like a meal or a mate. The amygdala gener- ates emotions and motivations based on the external sensory inputs of vision, hearing, and smell, rather than internal-envi- ronment inputs as with the hypothalamus (FIGURE 2.31a). The amygdala is best known for its role in the “aversive” emotion of fear, but it plays a role in linking external sensory inputs to positive emotional states as well. It is a quick learner of new emotional associations and a key site for emotional memory.
The hippocampus is another critically important site for memory and learning. A long, thin structure whose fanciful name means “seahorse,” it lies on the medial temporal lobes just behind the amygdala. It is traditionally considered a part of the limbic system, although its role in emotion and moti- vation is indirect. The hippocampus plays an important role in spatial navigation and episodic memory: memory for past personal experiences that occurred at a specific time and place, as opposed to memory for facts (FIGURE 2.31b). It also seems to be crucial for imagining future or hypothetical scenes, such as lying on a beach or going to a restaurant later in the day (Hassabis & Maguire, 2009; Schacter et al., 2012).
How are these functions of the hippocampus related to emotion and motivation? In a natural environment, the means of satisfying our needs are rarely nearby. A thirsty or- ganism probably will not be able to find water in the immedi- ate vicinity. Instead, it must search an internal map of its territory to locate a water source. Its internal map also needs to be kept up to date with past experiences, as new water sources appear and old ones dry up. Emotional associations
02-Eagleman_Chap02.indd 67 02/11/15 3:17 pm
68 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 68 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
with different locations are also important: being attacked at one water source might make that place worth avoiding in future, even if it happens to be in a convenient location. The hippocampus therefore has close connections with the
FIGURE 2.32 The ventricles of the brain.
(a) (b)
Central canal
Fourth ventricle
Cerebral aquaduct
Third ventricle
Third ventricle
Fourth ventricle
Cerebral aquaduct
Central canal
Interventricular foramen of Monro
Frontal (anterior) horn
* * *
*
*
AtriumBody Occipital (posterior) horn
Temporal (inferior) horn
Lateral ventricle
Lateral ventricles
amygdala, which helps attach emotional significance to places or events (R ichter-Levin & Akirav, 2000; Wells et al., 2011). It also has a major connection pathway via a loop called the fornix to a pair of nuclei in the hypothalamus, called the mamillary bodies. This circuit may be useful for linking the body’s current needs to the organism’s knowledge of places and past events (Sutherland & Rodriguez, 1989).
Parts of the cerebral cortex are sometimes considered limbic cortex. Limbic cortex forms a ring that starts with the visceral sensory cortex of the insula. The anterior part of the insula is next to the orbitofrontal cortex, which plays a vis- ceral motor role in generating internal bodily states, or so- matic markers, based on sensory input from the internal and external environment (Damasio, 1996). From here, the ring continues on to the medial wall and the anterior cingu- late cortex, which also performs visceral motor functions. It follows the cingulate gyrus over the corpus callosum to the posterior cingulate cortex, which plays a role in familiarity and emotional memory. It then gradually passes to the un- derside of the medial temporal lobe, whose functions mirror the navigation, memory, and emotional functions of the ad- jacent hippocampus and amygdala. The most anterior parts of the medial temporal lobe lie next to the anterior insula, completing the ring. This continuous ring of cortex in each hemisphere provides an interface between the internal world and all of the remaining sensory, motor, cognitive, and other functions of the cerebral cortex as a whole. In the chapters ahead, we will see how the limbic system plays an important role in many different types of cognitive functions.
The Ventricular System and Brain Function Our overview of the brain would be incomplete without de- scribing the critical role of the ventricular system to brain function. The four ventricles (cavities) in your brain are
FIGURE 2.31 The amygdala and hippocampus are key limbic structures. (a) The almond-shaped amygdala uses external sensory input to trigger emotional responses. (b) The hippocampus is a key structure for creating memories of specific locations in space and specific events in time.
Fornix
Mammilary bodies
Amygdala Hippocampus
Rapid evaluation of sensory input: Generation of emotional responses to external stimuli
Spatial navigation and episodic memory
(a) (b)
02-Eagleman_Chap02.indd 68 02/11/15 3:17 pm
Conclusion 69
# 158305 Cust: OUP Au: Eagleman Pg. No. 69 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
looping circuits from a broader point of view. We have seen short and simple circuits like reflex arcs, where signals move from sensory input to motor output in only a synapse or two. We have seen the more complex circuits of central pattern generators, command generators, the behavioral programs of the periaqueductal gray, and the still more elaborate ho- meostatic coordination of the hypothalamus. We have seen the reverberating and refining loops across multiple syn- apses through the cerebellum, thalamus, cortex, and basal ganglia.
How do we unravel this knot of knots? W hat is this whole system trying to accomplish, in the most general terms? To solve this problem, we’ll borrow a trick from ancient Greek mythology. The legendary hero Theseus used a ball of string to trace his path when he entered the original Labyrinth to battle the half-human Minotaur. A fter winning the battle, he followed the string through the Labyrinth’s twists and turns until at last he was safely back out of the maze.
Let’s flatten out all the complex anatomy of the brain into a two-dimensional maze. Now, let’s drop ourselves into a randomly chosen neuron in the brain and then follow its connections, synapse by synapse, until we emerge from the maze. W hat we’ll find is that, depending on which neuron we choose and whether we travel backward or forward, we’ll end up in one of four places.
We might end up in an external sensory neuron: a photo- receptor in the retina, an auditory receptor in the cochlea, a motion sensor in the vestibular organ, a touch receptor in the skin. External-world sensors make up one “edge” of the ner- vous system. Alternatively, we might end up at the end of a motor neuron, where it stimulates a skeletal muscle to move the body through the outside world. External-world move- ments make up a second edge of the ner vous system. Sometimes the paths between these two edges are short, as in a reflex arc. Others, through the cerebellum or cortex and basal ganglia, are long and looped in circles along the way. We’ll draw these edges as the upper two sides of a square; now we can see the many possible pathways of various differ- ent lengths (FIGURE 2.33).
We also might end up at an internal sensory neuron: a visceral pain receptor in the stomach, a pressure-sensing baroreceptor in a blood vessel, or an inflammation-sensing histamine receptor in the intestinal tract. These receptors make up a second sensory edge of the nervous system, with its own set of pathways distinct from the external sensory apparatus. Finally, we might end up traveling through the autonomic nervous system and its peripheral ganglia, to end up in a visceral motor neuron, stimulating an internal organ into its fight-or-flight or rest-and-regenerate mode. Again, the pathways between these two edges may be short, as in the simple visceral reflex arcs of the spinal cord. They may also be long, passing through multiple stages and loops, up to the hypothalamus, or even beyond, to the limbic regions of the cortex.
From this perspective, we can see the nervous system as an enormous network, half rooted in the body of the
filled with cerebrospinal fluid, not neurons (FIGURE 2.32). Two of these, the lateral ventricles, lie at the center of each hemisphere of the cerebral cortex, inside the white matter. These connect to the third ventricle, which lies along the midline of the brain, between the left and right thalamus. This connects to the fourth ventricle, a small triangular structure tucked between the brainstem and the cerebellum. The ventricles constantly produce cerebrospinal fluid, which circulates through the ventricles and over the surface of the brain and spinal cord. The fluid protects the brain from injury and helps to maintain a stable chemical environment for the neurons.
Conclusion We began this chapter by exploring the remarkable similari- ties between the brains of the hog-nose bat and the adult male sperm whale, and we learned how the brains of all mammals possess a common underlying structure. In fact, the cogni- tive abilities of human beings arise from a nervous system whose fundamental organization has been highly conserved over time.
In exploring the peripheral nervous system, we focused on its segmental organization as well as on its subdivisions into the somatic and autonomic nervous systems. The sym- pathetic and parasympathetic subsystems of the autonomic nervous system enable the body to operate in “fight-or- flight” and “rest-and-regenerate” modes, respectively. In ex- ploring the central nervous system, we learned about the structure and functions of the spinal cord as well as about the basics of how spinal reflexes and central pattern genera- tors work. Furthermore, we learned about the anatomy and the life-sustaining functions of the brainstem. We explored the complex structure of the cerebellum. Scientists agree that the cerebellum plays a key role in coordinating move- ment, but they debate its exact role in refining movements and in matching movements to the environment. Moreover, we learned about the role of the hypothalamus in homeosta- sis and the critical role of the thalamus as a relay station in the brain. Then we arrived at the highest level of the brain: the cerebral cortex. The cerebral cortex is divided into four lobes: the frontal lobe, temporal lobe, parietal lobe, and oc- cipital lobe, each of which contributes uniquely to brain function. We learned how scientists have studied the cytoar- chitecture of different cortical regions and have used the differences in circuitry across these regions to map the func- tional anatomy of the brain.
We learned about the structures of the basal ganglia, which help to initiate and maintain internally driven cortical activity. Finally, we concluded our journey through the ner- vous system by exploring the limbic system, which unites the internal and external worlds and is the system that mod- ulates motivation and emotions.
As we conclude our survey of the nervous system, let’s take a step back once again and look at all of these layers of
02-Eagleman_Chap02.indd 69 02/11/15 3:17 pm
70 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 70 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
External sensory pathway
Central
Peripheral
Levels of processing Levels of processing
Internal sensory pathway
External sensory input
Visceral motor output In
ter na
l s en
so ry
in pu
t
So ma
tic m
oto r o
ut pu
t
Somatic (external-world) motor pathway
Visceral (internal-world) motor pathway
FIGURE 2.33 The four-sided brain. We can think of the vast network of all neurons in the nervous system as having four edges. external-world sensory neurons for the major senses, internal-world sensory neurons for internal bodily senses, external-world motor neurons to move the body, and internal- world “motor” neurons to control the internal organs. The many layers of connections among these four edges make up the entirety of the central nervous system, allowing it to answer two key survival questions: “what’s going on inside and outside me?” and “what do I do about it?”
organism and half dangling out in the external environment in which it lives. Its pathways carry information back and forth, constantly trying to bridge the gap between these two worlds. To do so, they ask four questions, one for each edge of the network: W hat is happening out there? W hat is hap- pening in here? W hat should I do out there? W hat should I do in here?
In response to the need for more complex behaviors, the human ner vous system has added on longer and longer circuits, inserting more and more interneurons between the input and the output, and building more and more com- plex circuits. The result is a brain capable of meeting the body’s basic needs in astonishingly complex ways. As humans, we plan far ahead to sur vive and thrive. We delib- erately alter our environment over years and even entire generations to better suit our needs. We arrange complex societies in which many can prosper simply by acquiring and using specialized k nowledge. We pursue careers or chase lifelong dreams.
In this rarefied environment, many of us could live our entire lives and raise a new generation without ever needing to know how to forage for food, or avoid a predator, or survive a night in the cold. A ll of these activities are unprecedented
in the 3 billion–year history of life on Earth. As we have seen, the underlying structure of the human brain is not drasti- cally different from that of our nearest living relatives. In fact, we still share many common features with our ancestors of 5, 50, or even 500 million years ago. Yet there is clearly something new and unique about the way that human beings are using the basic machinery of cognition. As we look at this machinery over the rest of the chapters in this book, try to keep this point in mind. Neuroscience is ultimately a con- certed, precise effort to understand ourselves using the very brains we study. Neuroscience should tell us a story not only about our nervous system, but also about human nature: who we are and what we are capable of achieving.
02-Eagleman_Chap02.indd 70 02/11/15 3:17 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 71 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
KEY PRINCIPLES
organized as folia, grouped into lobules, which are further grouped into lobes.
• The extensive circuitry of the cerebellum allows for smooth, accurate, coordinated movements.
• The hypothalamus coordinates homeostatic func- tions, including sleep and eating, to keep the body’s internal environment in balance. The thalamus is a relay center that coordinates the flow of sensory information to the cerebral cortex. It also coordi- nates information flow between distant areas of the cerebral cortex itself.
• The cerebral cortex provides the nervous system’s most elaborate circuitry for sensory, motor, and intermediate functions. It is divided into four lobes: the frontal lobe, temporal lobe, parietal lobe, and occipital lobe.
• Circuits of the basal ganglia initiate and maintain internally driven cortical activity, particularly that related to motor control.
• The limbic system includes the hypothalamus, parts of the thalamus, the substantia nigra, the amygdala, the hippocampus, and the limbic re- gions of the cortex. Collectively, these regions have important motivational and emotional functions.
• Overall, the circuits of the nervous system bridge the gap between sensation and action in the inter- nal and external worlds, linking the questions of What’s happening in here and out there? to What should I do in here and out there?
• All vertebrates have a nervous system featuring segmental organization and an expansion at the front end for centralized control, using internal and external sensory inputs. Ever-more-complex circuits linking sensory input and motor output are added on top of one another, at multiple levels of the nervous system, to allow more and more com- plex forms of behavior.
• The peripheral nervous system collects sensory input from both inside and outside the body and transmits it to the central nervous system. The pe- ripheral nervous system carries output signals to the internal organs as well as the muscles of the body.
• Simple spinal reflexes allow sensory inputs to direct motor outputs with minimal involvement of the central nervous system. Central pattern gen- erators allow for more complex, coordinated movements such as locomotion.
• In the brainstem, more elaborate reflexes and central pattern generators operate in the handling of special sensory input and in the control of the special movements of the head region.
• The brainstem includes the medulla oblongata, pons, and midbrain. These regions relay sensory and motor information between the brain and spinal cord and are the point of origin for most of the cranial nerves.
• The cerebellum contains the majority of neurons in the central nervous system, and these are
KEY TERMS
An Overview of the Nervous System spinal cord (p. 39) central nervous system (p. 41) peripheral nervous
system (p. 41) forebrain (p. 41) midbrain (mesencephalon)
(p. 41) hindbrain (p. 41)
telencephalon (p. 41) diencephalon (p. 41) metencephalon (p. 41) myelencephalon (p. 41) rostral (p. 42) caudal (p. 42) dorsal (p. 42) ventral (p. 42) anterior (p. 42)
posterior (p. 42) superior (p. 42) inferior (p. 42) medial (p. 42) lateral (p. 42) ipsilateral (p. 42) contralateral (p. 43) distal (p. 43) proximal (p. 43)
Key Terms 71
02-Eagleman_Chap02.indd 71 02/11/15 3:17 pm
72 PART 1 • ChAPTER 2 The Brain and Nervous System
# 158305 Cust: OUP Au: Eagleman Pg. No. 72 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The Peripheral Nervous System sensory neurons (p. 43) motor neurons (p. 43) neuromuscular junction (p. 43) neurotransmitters (p. 43) somatic nervous system (p. 43) autonomic nervous
system (p. 43) sympathetic nervous
system (p. 43) parasympathetic nervous
system (p. 43) dermatomes (p. 45)
The Spinal Cord gray matter (p. 47) white matter (p. 47) stem cells (p. 47) ventral horns (p. 48) dorsal root ganglion (p. 48) reflex (p. 48) synapse (p. 48) interneuron (p. 48) stretch reflex (p. 48) central pattern generator
(p. 49)
The Brainstem brainstem (p. 51) medulla oblongata (p. 51) pons (p. 51) cranial nerves (p. 51) cerebrum (p. 51) nuclei (p. 52) superior colliculus (p. 53) inferior colliculus (p. 53) periaqueductal gray matter
(p. 53)
reticular formation (p. 54) locus coeruleus (p. 54) norepinephrine (p. 54) substantia nigra (p. 54) dopamine (p. 54) midbrain raphe nuclei
(p. 54) serotonin (p. 54)
The Cerebellum cerebellum (p. 56) Purkinje cells (p. 56) forward model (p. 57)
The Diencephalon: Hypothalamus and Thalamus
hypothalamus (p. 57) thalamus (p. 57) set points (p. 57) hormone (p. 58) pituitary gland (p. 58) oxytocin (p. 58) cerebral cortex (p. 58) lateral geniculate nucleus
(p. 59) photoreceptors (p. 59) primary visual cortex
(V1) (p. 59) association areas (p. 59) reticular nucleus (p. 60) deep brain stimulation (p. 61)
The Telencephalon: Cerebral Cortex and Basal Ganglia
gyri (singular, gyrus) (p. 63) sulci (singular, sulcus) (p. 63) corpus callosum (p. 63) frontal lobe (p. 63) temporal lobe (p. 63)
parietal lobe (p. 63) occipital lobe (p. 63) central sulcus (p. 63) precentral gyrus (p. 63) primary motor cortex (p. 63) prefrontal cortex (p. 63) orbitofrontal cortex (p. 63) postcentral gyrus (p. 63) primary somatosensory
cortex (S1) (p. 63) precuneus (p. 63) cingulate gyrus (p. 64) superior temporal gyrus
(p. 64) primary auditory cortex (A1)
(p. 64) fusiform gyrus (p. 64) insula (p. 64) basal ganglia (p. 64) caudate nucleus (p. 64) putamen (p. 65) nucleus accumbens (p. 65) globus pallidus (p. 65) subthalamic nucleus (p. 65)
Uniting the Inside and Outside Worlds
limbic system (p. 66) amygdala (p. 67) hippocampus (p. 67) episodic memory (p. 67) fornix (p. 68) mamillary bodies (p. 68) somatic markers (p. 68) anterior cingulate cortex
(p. 68) ventricles (p. 68) cerebrospinal fluid (p. 69)
REVIEW QUESTIONS
1. Why do vertebrates collect their neurons into a brain?
2. Draw a simple human body as seen from the front and the side. A stick figure will do. Now add labeled arrows illustrating all 14 anatomi- cal directions listed in this chapter. Draw and
label three lines showing the angles of an axial, a coronal, and a midsagittal slice through the head.
3. List the components and subcomponents of the peripheral nervous system. Summarize the function of each one in a short sentence.
02-Eagleman_Chap02.indd 72 02/11/15 3:17 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 73 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
4. Define dermatomes and myotomes. What are the likely effects of a spinal cord injury at the level of the neck? At the level of the lower back?
5. Explain the mechanism by which a tap below the kneecap can produce a vigorous kick of the leg. Explain how spinal cord circuits can generate simple, alternating movements useful for locomotion.
6. Describe three important functions of the brain- stem. Why can an injury to the medulla oblon- gata be fatal, even if the rest of the brain is unharmed? Describe three categories of behav- ior coordinated in the midbrain.
7. Describe the homeostatic functions of the hypo- thalamus. How might your hypothalamus try to
compensate when it senses that you are dehydrated?
8. List the four lobes of the cerebral cortex, and give one example of a function contributed by each lobe. What is meant by the term cytoarchi- tecture, and why is cytoarchitecture helpful in understanding brain function?
9. What are the structures of the basal ganglia? How do they contribute to the activity of the ce- rebral cortex?
10. What is meant by the term limbic system? What brain structures are typically considered part of the limbic system? What functions do they perform?
CRITICAL-THINKING QUESTIONS
1. Why do you think that the sympathetic and para- sympathetic nervous systems exist in com- pletely separate segmental regions? What functional advantage(s) do you think such seg- mentation might convey?
2. Imagine that, for one week, you had to suffer from the malfunction of one of three brain structures: your cerebellum, hippocampus, or amygdala. During that week, you had to try to go about your daily routine as much as possible. Which of these three brain structures would you choose to malfunction, and why? In your
response, consider the neural pathways in which each of these three brain structures is involved.
3. Imagine that you could choose to have the func- tioning of one of the four lobes of your cerebral cortex significantly enhanced: the frontal lobe, temporal lobe, parietal lobe, or occipital lobe. Which lobe would you choose, and why? In what specific ways might your everyday life be differ- ent if that lobe of your brain were even more powerful in its functioning than it is now?
Critical-Thinking Questions 73
02-Eagleman_Chap02.indd 73 02/11/15 3:17 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 74 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
74
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Distinguish the major types of cells in the brain.
• Isolate the structures of the neurons that allow neurons to collect, integrate, and output signals.
• Explain how chemical signaling operates at a synapse.
• Describe the mechanisms of an action potential.
• Summarize the type of information carried by action potentials.
• Characterize how the neural code can be distributed across populations of neurons.
03-Eagleman_Chap03.indd 74 02/11/15 3:21 pm
75
# 158305 Cust: OUP Au: Eagleman Pg. No. 75 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Neurons and Synapses
STARTING OUT: The Kabuki Actor and the Pufferfish
The Cells of the Brain
RESEARCH METHODS: Visualizing Neurons and Their Products
Synaptic Transmission: Chemical Signaling in the Brain
THE BIGGER PICTURE: Psychoactive Drugs Spikes: Electrical Signaling in the Brain
CASE STUDY: Multiple Sclerosis
NEUROSCIENCE OF EVERYDAY LIFE: The Magic of a Local Anesthetic
What Do Spikes Mean? The Neural Code
RESEARCH METHODS: Recording Action Potentials with Electrodes
Individuals and Populations
CHAPTER 3
03-Eagleman_Chap03.indd 75 02/11/15 3:21 pm
76 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 76 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
stream about what is being sensed and what the body should do about it. Complexity emerges as the population of neu- rons between motor and sensory systems expands to enable more sophisticated prediction. And at the heart of it all is an unimaginably vast network of cells within the nervous system that send lightning-fast signals through a network to allow this all to happen. In this chapter we’ll study those cells in detail.
In the previous chapter we learned how the overall struc- ture of the brain allows signals from the periphery (for example, from your eyes and fingertips) to flow in, get processed, and go out to influence in the outside world (for
The overall goal of the nervous system is to enable the organ- ism to move its body appropriately to succeed at the four F’s: feeding, fleeing, fighting, and reproducing. To move adap- tively, the organism uses information about what is going on inside its body (I am hot; I am thirsty) and outside its body (there is shade under that ledge; there is water at the end of the meadow), so that it can make predictions about external and internal events and act accordingly. Your eyes and ears carry signals about external events by responding to physical stimuli such as light patterns and sound waves. At the behav- ioral end, motor neurons command the muscles to contract or relax, and this loop constructs an ever-changing decision
STARTING OUT: The Kabuki Actor and the Pufferfish Bandō Mitsugorō VIII (1906–1975) was one of Japan’s most beloved Kabuki actors and was named as a living treasure during his lifetime by the government. In a drunk and pompous display in a restaurant, he claimed immunity from poisoning and demanded four livers of the infamous puffer fish, known in Japanese as fugu fish (FIGURE 3.1). The chef, who later claimed to be unable to turn down the request of such a famous celebrity, apprehensively ser ved the livers. Mitsugorō bravely con- sumed the livers and quickly went into complete paralysis. Since he could not breathe without proper muscle control, he died after seven hours.
Fugu fish contains a molecule called tetrodotoxin, an extremely powerful poison with no known antidote. There is nothing espe- cially striking about the structure of the molecule itself: like all mol- ecules, it is simply several atoms hooked together in a particular way. What makes it lethal?
The answer is that tetrodotoxin, like many similar poisons, prevents the transmission of action poten- tials, the electrical signals by which neurons communicate quickly over long distances. In the presence of tetrodotoxin, the brain sends out commands to which the body cannot respond, because the electrical sig- nals never reach their destination. Since the muscles cannot move
(and this includes the diaphragm muscles critical to breathing), vic- tims die of suffocation.
In this chapter we will learn about the processes with which te- trodotoxin interferes. We will learn how neurons communicate with one another, what their signals mean, and why getting drunk and claiming that you cannot be poisoned does not accord with your neurobiology.
FIGURE 3.1 The tasty but deadly fugu, or puffer, fish contains tetrodotoxin, which prevents the transmission of action potentials.
03-Eagleman_Chap03.indd 76 02/11/15 3:21 pm
The Cells of the Brain 77
# 158305 Cust: OUP Au: Eagleman Pg. No. 77 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
example, picking up a cup of coffee). But that level of under- standing, by itself, does not allow us to understand how pharmaceutical drugs work, why drugs of abuse are self- administered, what happens in diseases like A lzheimer’s, how hormones influence your drives, and why people make risky decisions—all topics that we’ll be tackling in later chapters. To set the stage for these issues, we’ll now zoom in to the individual cells in the nervous system and how they communicate with each other. This chapter will contain many terms that may be new to you, but try to keep in mind a key principle: neurons communicate in a vast network by chemical and electrical signals. We need to appreciate the detailed biology to understand how the signaling carries in- formation and how that information can be modified by pharmaceutical drugs, hormones, drugs of abuse, and disease.
The Cells of the Brain Imagine how you would think of the brain if you lived 120 years ago. Because the brain has the consistency of slightly hardened mashed potatoes, you would have no idea of what it was actually composed. You might suppose, for
example, that it is made of a continuous series of tubes through which substances flow. And you’d be in good com- pany. Just over a century ago, before the advent of good mi- croscopy, most scientists supposed that neural tissue was a continuous network like the blood vessels (Cimino, 1999; Lopez- Munoz, Boya, & A lamo, 2006). This idea was over- turned by the Spanish neuroscientist Santiago Ramón y Cajal, who used new techniques to stain and visualize brain tissue and realized that the brain is built of billions of discrete cells (FIGURE 3.2). His “neuron doctrine,” which stated that cells in the brain are separate entities, ushered in an impor- tant new mystery: the cells somehow need ways to commu- nicate with each other over the tiny spaces that separate them. This chapter is about that signaling. Amazingly, from the enormous symphony of activity of these cells, behavior and cognition emerge.
Neurons: A Close-Up View The most important ty pe of cell in your ner vous system is a neuron (FIGURE 3.3). The human brain contains almost 100 billion neurons (Herculano-Houzel, 2012). Neurons are, in most ways, like all the other cells in your body: they have a membrane, nucleus, and specialized organelles and they produce, traffic, and secrete chemicals. The cell ’s membrane separates the cell ’s components from the envi- ronment outside of the cell. Like other cell ty pes, neurons have proteins that are inserted into their membranes, and these proteins allow the cell to interact with its outside en- vironment. But neurons possess an additional property that distinguishes them: because of the particular pro- teins on their surfaces, they can transmit electrical signals quick ly over long distances. A nd when those electrical sig- nals arrive at their end point(s), they trigger a specialized form of chemical signaling. In the next few sections we will learn how these forms of signaling take place. But first we will turn to the specialized anatomy that gives a neuron its remarkable capabilities.
Neurons have four zones of importance. The first con- sists of the dendrites, which are long, branching extensions from the cell body. Dendritic trees can take on many shapes and sizes—from single branches to large cones, pancakes, or
Nucleus
Dendrites
Soma (cell body)
Axon
Terminal branches
Axon terminals
FIGURE 3.3 A typical neuron in the cortex.
(b)(a)
FIGURE 3.2 Visualizing brain cells. (a) Santiago Ramón y Cajal (1852–1934). (b) One of Cajal’s first renderings of a neuron impregnated with his staining technique.
03-Eagleman_Chap03.indd 77 02/11/15 3:22 pm
78 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 78 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
(a)
(b)
Soma (cell body)
Nucleus
Nucleolus
Axon hillock
Neuron soma
Dendritic spine
Dendrites
Dendrites (receptive regions)
Nissl bodies
Myelin sheath
Terminal branches
Axon terminals
Node of Ranvier
or cell body (soma is Greek for “ body”) (FIGURE 3.5). The key feature of the soma is the cell’s nucleus, which is the control center of the cell that regulates cell activity, including gene expression. A typical soma spans about 10–25 micrometers, although the sizes can vary widely across different neuron types. As we will see in this chapter, the soma plays a key role in integrating (that is, summing up) the signals coming in from the dendrites.
Emerging from the soma is the single, long slender process of the a xon, or ner ve f iber, which is the third zone of impor tance (FIGURE 3.6). The a xon is an ex tension that reaches long distances beyond the soma, and it is es- sentially a cable to conduct signals rapidly across long dis- tances, as we w ill see shor tly. The a xon dif fers from the dendrites in three ways. First, there is only one a xon coming from a neuron, whereas there can be many den- dritic ex tensions. Second, a xons tend to remain constant in diameter all along their leng th, whereas dendrites are tapered. Finally, a xons tend to be much longer than den- drites: dendritic trees rarely ex tend more than 3 millime- ters, whereas a xons carr y ing signals from your spinal cord to your big toe (the sciatic ner ve) r un the entire leng th of your leg—and in giraf fes, a xons several meters in leng th r un the entire span of the neck!
A typical axon will branch robustly at its end, typically splitting into about 10,000 axon terminals (sometimes called axonal boutons or buttons)—and these terminals constitute the fourth zone of importance (FIGURE 3.7). The terminals are identifiable as small swellings at the end tips and, as we will see, they contain packages of chemicals that can be released into the space between cells. The terminals, therefore, are optimized for the output of signals.
A xon terminals are t y pically found in close prox imit y to the dendrites and somas of other cells, and such junc- tions are called sy napses (syn is Greek for “together” and haptein “to clasp”). Santiago R amón y Cajal poetically
described the sy napse as a “protoplasmic k iss” bet ween t wo cells (R amón y Cajal, 1937). The main location of signal transmission, the sy napse links an a xon to other neurons (in the central ner vous
system) or to a neuron, muscle, or gland (in the peripheral ner vous system). A lthough
most sy napses occur at the a xon ter- minals, they can also ex ist along the a xon itself, and in this case they are k now n as en passant sy napses. The t y pical sy napse connects an a xon to a
FIGURE 3.4 Dendrites. The integrators of thousands of tiny chemical signals come in a variety of shapes.
FIGURE 3.5 The cell body, or soma, is the central command center of a neuron. The dendrites and a single axon grow from the soma, the former for collecting incoming signals and the latter for transmitting outgoing signals over long distances.
spheres (FIGURE 3.4). As we will explore in the next section, dendrites are specialized for collecting information from thousands of tiny chemical signals that they receive all along their extent.
By responding to chemical messages along their intricate branching patterns, dendrites collect a great deal of informa- tion and pass it to the second zone of importance: the soma
03-Eagleman_Chap03.indd 78 02/11/15 3:22 pm
The Cells of the Brain 79
# 158305 Cust: OUP Au: Eagleman Pg. No. 79 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Dendrites: Collects thousands of incoming signals
Soma: Integrates signals from dendrites; generates signal that will travel down axon
Axon: Conducts signals rapidly across long distances
FIGURE 3.6 An axon is a single, slender extension from the soma. It is essentially a cable to conduct signals rapidly across long distances.
Many Different Types of Neurons Dendrites, soma, axons, and axon terminals are characteris- tic features of all neurons, but there is nonetheless a great deal of diversity among neuron types.
The most common way to classif y neurons is by their function. Sensory neurons are those that directly respond to signals from the outside environment—for example, light, sound waves, pressure, or odors (FIGURE 3.8a). Motor neurons have direct output to muscles or glands; they are the final step for signals to exit the nervous system and effect change in the body or environment (FIGURE 3.8b). You may sometimes hear the term afferent neuron for the incoming (sensory) neurons and efferent neuron for the outgoing (or motor) neurons. (These terms can be remembered by asso- ciating afferent with arrival and efferent with exit). In mam- mals the vast majority of neurons cannot be classified as either sensory or motor—instead, they are the interneurons between the sensation of a signal at the one end and the action at the other end (FIGURE 3.8c). Primitive animals such
dendrite or soma. T here are also sy napses that can join a xon to a xon or dendrite to dendrite, although these are more rare.
The number of synapses in the brain is boggling: in a three-year-old brain, the total number of synapses is esti- mated to be a quadrillion (1 followed by 15 zeros) (A lonso- Nanclares, Gonzalez-Soriano, Rodriguez, & DeFelipe, 2008). Synapse numbers decrease with age through a natu- ral pruning process; therefore, an adult brain contains some- where between 100 and 500 trillion synapses (Drachman, 2005) in a total volume of about 1,200 cubic centimeters (Cosgrove, Mazure, & Staley, 2007). For a typical college student, this means that a cubic millimeter of cerebral cortex contains several billion of these tiny connections—about as many people as exist on the planet.
As we will come to understand in this chapter, the four zones of a neuron foreshadow its critical functions: collecting (dendrites), integrating (soma), conducting (axon), and outputting information (axon terminals). The synapses are the point where the axon terminals contact the next cells, and here the chain of signaling continues.
Neurotransmitter molecules
Dendrites
Electrical impulses
Axon
Neuron
Synapse
Receptor
FIGURE 3.7 Axon terminals are the end points of the axon, where chemical signals are released.
03-Eagleman_Chap03.indd 79 02/11/15 3:23 pm
80 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 80 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
brain, almost all the neurons are “in between” the sensory and motor ends.
A n alternative way of classif y ing neurons hinges on their basic shape. In this classification scheme, multipolar neurons are those w ith multiple dendrites (FIGURE 3.9a); these are the most common class. Bipolar neurons, on the other hand, are composed of a single dendrite on one end and a single a xon on the other (FIGURE 3.9b); these are often found in sensor y neurons such as the retina and inner ear. Finally, monopolar neurons have only a single extension that leaves the soma and branches in t wo directions (FIGURE 3.9c). One end of a monopolar neuron receives the information and the other end ser ves for output. This t y pe of neuron is t y pically found in sensor y neurons that signal touch and pain.
A lthough there are many types of neurons, it should be noted that they all share in common the feature of being postmitotic—that is, they do not divide like many other cell types in the body.
Glial Cells Neurons receive the most scientific attention because of their long reach and ability to carry rapid electrical signals. But there is another type of cell that plays a wide range of supporting roles in the nervous system: the glial cells, or glia. “Glia” comes from the Greek word for “glue,” reflecting the original idea that they were only meant to hold the network of neurons together (Kettenmann & Verkhratsky, 2008). A lthough the full functional capacity of the glial cell is still under intensive study, it is clear that these cells play several roles, providing ways to speed up the signaling from neurons, regulating the concentrations of extracellu- lar chemicals, and determining the extent to which net- works of neurons can modify their connections (A llen & Barres, 2009).
A lthough neurons come in hundreds of forms, glia come in only four basic types. The first is oligodendrocytes, large cells whose main function is to wrap a layer of “insulation” around axons—a process known as myelination—similar to the way that a copper wire is wrapped in rubber. The conse- quence of myelination is the speeding up of electrical signal- ing by neurons, a topic we’ll explore in detail later in the
Dendrites Dendrite Axon
SomaSomaSoma
Branches
Multipolar neurons have multiple dendrites.
Bipolar neurons are composed of a single dendrite on one end, and a single axon on the other.
Monopolar neurons have only a single extension that leaves the soma and branches in two directions.
(a) (b) (c)
FIGURE 3.9 Classifying neurons by their shape. Examples of (a) multipolar neurons, (b) bipolar neurons, and (c) monopolar neurons.
(b) Motor neuron
(a) Sensory neuron
(c) Interneurons
MuscleMyelin sheath
Myelin sheath
Receptor cell
FIGURE 3.8 Different types of neurons. Examples of (a) sensory neurons, (b) motor neurons, and (c) interneurons. Interneurons can be of two types: those with long projections to other regions are termed projection interneurons, whereas those that stay within a region are termed local interneurons.
as jelly fish have neurons that contain both sensory and motor qualities in the same cell, but this combination is not found in more advanced species. Evolutionarily, it is thought that mammalian brains have developed by a dissociation of sensory and motor neurons into separate cell types, fol- lowed by the gradual insertion of more and more neurons in between (Miller, 2009). By the time we reach the human
03-Eagleman_Chap03.indd 80 02/11/15 3:23 pm
The Cells of the Brain 81
# 158305 Cust: OUP Au: Eagleman Pg. No. 81 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
RESEARCH METHODS: Visualizing Neurons and Their Products The discovery that neurons are dis- crete, fundamental units of the ner- vous system was made possible by the ability to stain them. Several techniques allow the visualization of neurons. Golgi staining is a tech- nique that impregnates some frac- tion of neurons with a dark material, allowing the entirety of individual cells to be seen under a microscope (FIGURE 3.10a). This is the method that birthed Ramón y Cajal’s neuron doctrine. Another technique, Nissl staining, uses a chemical that binds to the RNA in cell bodies, thereby allowing the visualization of somas (FIGURE 3.10b). Nissl staining is most commonly used for judging sizes of cells and their densities.
Several other methods are utilized to obtain detailed pictures of nervous tissue. In the technique of autoradi- ography, a radioactive substance is designed to be taken up by specific cells but not by others (FIGURE 3.10c). Then, when a photographic emulsion is placed over thin slices of the brain tissue, the emulsion is exposed by the radioactivity in the same way that film is exposed by light. In this way, it can be seen which cell types absorbed the substance in question (for example, a pharmaceutical drug).
In the technique of immunocyto- chemistry, antibodies are developed that bind only to specific proteins (FIGURE 3.10d). These antibodies are washed onto a slice of brain tissue, and they attach wherever the protein of interest is being expressed. With some chemical steps, these antibod- ies can be visualized, revealing the
exact locations of the protein within the cell. A related technique is to use radioactively labeled stretches of RNA or DNA that will bind to specific
stretches of messenger RNA (mRNA); this is called in situ hybridization, and it reveals which cells have expressed a gene of interest (FIGURE 3.10e).
A
C
FIGURE 3.10 Different techniques to bring the invisibly small world of neurons to light. (a) Golgi staining, (b) Nissl staining, (c) autoradiography, (d) immunocytochemistry, and (e) in situ hybridization.
(a) (b)
(d)
(e)(c)
03-Eagleman_Chap03.indd 81 02/11/15 3:23 pm
82 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 82 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
chapter. A single oligodendrocyte wraps the axons of up to 50 different neurons (FIGURE 3.11a). Oligodendrocytes are found only in the central nervous system; the function of myelination is accomplished in the peripheral nervous system by a second type of glial cell, the Schwann cells (FIGURE 3.11b). Schwann cells are quite similar in function to oligodendrocytes, with the minor exception that a Schwann cell wraps myelin around only a single axon (Bhatheja & Field, 2006). Myelin sheaths are not continuous along the length of an axon, but instead come in short segments (about 100 micrometers), appearing in a microscope like a string of sausages. The gaps between the myelinated segments are called nodes of Ranvier, which you will learn more about later in this chapter. Given the length of axons in the periph- eral nervous system (recall the sciatic nerve that runs the length of your leg), this translates to an enormous number of glial cells that contribute to the insulation of each axon. Note that many axons do not become myelinated, especially those in the cortex, but most subcortical and peripheral axons do take on myelination.
The third type of glial cell is the astrocyte, named for its star shape (FIGURE 3.12a; astro from the Greek “star,” cyte from “cell”). Beyond physical structural support, astrocytes per- form critical functions in maintaining the balance of chemi- cals outside the neurons, the repair of injury in the central nervous system, the contribution of nutrients, the regula- tion of local blood flow to a region, and the release of chemi- cal signals (Fellin, 2009). The fourth type of glial cell is the microglia (FIGURE 3.12b). Making up 20% of the glial cell population, these small cells are the front line of immune FIGURE 3.12 Two more types of glial cells. (a) Astrocytes and (b) microglia.
(a) (b)
defense in the central nervous system: they are constantly on the move, searching for any infectious agents that might damage normal neural tissue. W hen they detect a foreign body, they consume and destroy it to prevent disease and in- flammation (K reutzberg, 1995).
W hen you think about the pieces and parts that make up the brain, don’t forget to consider the amazing quality of the big picture: an average desktop computer makes bil- lions of computations in a single second. By contrast, neu- rons in the cerebral cortex, on average, fire only about 7 times per second and rarely exceed firing rates of 50 times per second (O’Connor, Peron, Huber, & Svoboda, 2010). How can the brain perform feats effortlessly that our most advanced supercomputers cannot, such as finding a path through a forest, recognizing the face of a loved one, or
(a)
Axon
Myelin lamellae
Myelin sheath
Oligodendrocyte Axon
Axon
(b) (c)
Schwann cell
FIGURE 3.11 Some glial cells myelinate axons. (a) In the central nervous system, a single oligodendrocyte will wrap up to 50 different axons with myelin sheaths. (b) In the peripheral nervous system, myelination is accomplished by Schwann cells, which wrap around a single axon. Note that the layer of insulation is not continuous, but exists in small sections. (c) Transmission electron micrograph of a myelin sheath.
03-Eagleman_Chap03.indd 82 02/11/15 3:23 pm
Synaptic Transmission: Chemical Signaling in the Brain 83
# 158305 Cust: OUP Au: Eagleman Pg. No. 83 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
the heart) would cause the heart to slow down. Loewi sus- pected the final mechanism of action was not electrical, but chemical in nature. So he isolated a frog’s heart in a bath of salt solution, stimulated the nerve, and then extracted the fluid of the bath. He reasoned that if a chemical had been re- leased by the nerve stimulation, it should still be present in the bath. And indeed, when he pumped the bath solution onto a second frog heart, it immediately slowed the heart rate. He correctly concluded that the signal from the axon was chemical (Zigmond, 1999).
Loewi had just discovered neurotransmission, the trick by which cells of the nervous system communicate across small gulfs of space to each other or to targets such as the muscle cells of the heart. (Loewi was awarded the Nobel Prize in 1936 for this breakthrough.) The released chemi- cals are called neurotransmitters. The neurotransmitter is released by the presynaptic cell and, by diffusing from its point of release, is felt as a change of chemical concentration at the postsynaptic target (FIGURE 3.14). Note that, in this case, the signal transmission is one way, carrying a signal from the axon to the dendrite, but not the other way around. The synaptic cleft—the little space between the pre- and postsynaptic cells—is a mere 20–50 nanometers (billionths of a meter) across, and this small distance allows the con- centration of neurotransmitter to rise and decay rapidly. Just inside the membrane of a presynaptic cell, the neu- rotransmitter molecules are packaged inside small spherical packages called synaptic vesicles. The release of the neu- rotransmitter into the extracellular space occurs when the vesicle fuses with the outer membrane and the molecules spill out into the cleft.
Before we explore what happens when the molecules reach the postsynaptic target, it is important to know that many different chemicals function as neurotransmitters; we will briefly survey these now.
writing a poem? The answer seems to lie in the brain’s mas- sive parallelism: the neurons and glial cells may be slow, but they are massive in quantity. From the parallel interaction of hundreds of billions of these cells, the magic of cognition and behavior emerges. We now turn to details of that interaction.
Synaptic Transmission: Chemical Signaling in the Brain In the previous section you learned that dendrites are the first place that signals are received by a neuron. We now turn to this in more detail: what do those signals look like, and how are they collected? In this section, we will describe two of the four critical functions of the neuron that we men- tioned before: outputting and collecting information. In later sections, we’ll return to the other two functions, inte- grating and conducting information.
Release of Neurotransmitter at the Synapse How do neurons communicate across the small spaces that divide them? This was recognized as a problem in the early 1900s, but it was not solved until the 1920s, when a scientist named Otto Loewi performed a simple and elegant experi- ment (FIGURE 3.13). It was known that electrically stimulating a particular nerve in the frog (the vagus nerve that leads to
(b)(a)
Saline pumped
from heart 1 to
heart 2
Heart 1 Heart 2
Stimulation
Heart rate
Vagus nerve Stimulator
Stimulating the vagus nerve causes heart 1 to slow…
…pumping the saline bath causes heart 2 to slow as well.
FIGURE 3.13 (a) Otto Loewi and (b) his experimental design that led to the discovery of neurotransmission.
03-Eagleman_Chap03.indd 83 02/11/15 3:23 pm
84 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 84 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
in the peripheral nervous system). Aspartate is another ex- citatory amino acid neurotransmitter, whereas GA BA (gamma-aminobutyric acid) and glycine are common in- hibitory neurotransmitters.
Beyond these common neurotransmitters, there are also peptide neurotransmitters (peptides are short strings of amino acids). Examples of these “neuropeptides” include cholecystokinin, somatostatin, and neuropeptide Y; more than 50 have been found to date.
Finally, signals can also be carried by gases such as nitric oxide and carbon monoxide, which diffuse directly through cell membranes to effect changes at neighboring locations. These soluble gases often work in the opposite way from what we described before, being produced in the dendrites of one cell and crossing the synapse backward, to affect the axons of the presynaptic cell. For that reason, they are often referred to as retrograde transmitters. TABLE 3.1 summa- rizes the categories of neurotransmitters.
Most neurons release one peptide neurotransmitter in addition to one, or a few, of the smaller neurotransmitters (acetylcholine, monoamines, and amino acids). Generally, however, neurons will release the smaller neurotransmitters first and only start to release the larger peptide neurotrans- mitters after they have been stimulated repeatedly.
Given the long litany of neurotransmitter types, the im- portant point to appreciate is the variety of substances that all serve the function of carrying information between neu- rons. Through evolutionary timescales, neurons have forged partnerships of presynaptic secretion and postsynaptic detection using chemicals of all types. We now turn to the details of the postsynaptic detection.
Types of Neurotransmitters The neurotransmitter discovered in Loewi’s experiment was a molecule called acetylcholine. This molecule serves as an excitatory neurotransmitter in the peripheral nervous system, causing muscle contractions when released at the junction between the nervous system and the muscular system. Since Loewi’s day, dozens more neurotransmitters have been dis- covered. Because these neurotransmitters will come up again throughout the book, we will briefly cover the different cate- gories—this will be an integral part of everything that comes later. Don’t get lost in the details; for now just try to appreci- ate that there are many different types of molecules that have been capitalized on by neurons to carry signals.
Acetylcholine, the first neurotransmitter discovered, is the only one that stands in a category of its own. Several other neurotransmitters fall under the categor y of the monoamines: examples are dopamine, epinephrine, norepi- nephrine (all known as catecholamines), serotonin, and melatonin. Dopamine, as an example, serves as the critical information-carrying molecule in the brain’s reward systems and is the target of drugs of addiction such as cocaine and amphetamines (Chapter 14). It is also the main neurotrans- mitter implicated in schizophrenia (Chapter 16).
A lthough the monoamines are large molecules, it is also possible to carry signals with tiny molecules such as amino acids, the building blocks of proteins—and these thus com- prise a third category of neurotransmitter. The amino acid neurotransmitter glutamate is the most common excitatory transmitter in the central nervous system (as acetylcholine is
(a)
(b)
Vesicle
Synaptic cleftTransporter
ReceptorsNeurotransmitter
Dendrite
Action potential
Presynaptic neuron
Postsynaptic neuron
Axon
FIGURE 3.14 Vesicles carrying neurotransmitter molecules dock with the presynaptic membrane, releasing the signaling molecules into the synaptic cleft. The neurotransmitters diffuse across the cleft and interact with receptors on the postsynaptic target.
03-Eagleman_Chap03.indd 84 02/11/15 3:24 pm
Synaptic Transmission: Chemical Signaling in the Brain 85
# 158305 Cust: OUP Au: Eagleman Pg. No. 85 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
proteins inside the cell. These in turn relay, amplify, and transform the signal. Because so many receptor types are G-protein coupled, this allows the cell to develop sophis- ticated signaling cascades that integrate several signals from the outside. The second messengers triggered by metabo- tropic receptors can serve many different functions, modu- lating the activity of neighboring ion channels, activating or deactivating enzymes within the cell, or changing which genes are expressed within the cell. These metabotropic receptors are a large and important family of receptors, so im- portant, in fact, that about half of all medical drugs target these receptors. The discovery of these receptors was awarded the 2012 Nobel Prize in Chemistry.
The effects of metabotropic receptors tend to operate on a much slower time scale than ionotropic receptors. For any given neurotransmitter, such as serotonin, there can be literally dozens of different subtypes of ionotropic or metabotropic receptors found in various parts of the nervous system. Thanks to this immense variety, a single neurotrans- mitter can serve many different types of functions through- out the body.
W hen neurotransmitters bind to receptors, they do not remain there for long. Instead, the binding is brief, and their presence in the synaptic cleft is quickly and precisely cleaned up. This clean-up occurs by one of three mechanisms: degradation, in which the neurotransmitter molecule is broken apart by other molecules; diffusion, in which the neu- rotransmitter moves out of the synapse, down its chemical concentration gradient; or reuptake, in which specialized protein transporters in the membrane will selectively pull the neurotransmitter back inside the cell, presynaptically, postsynaptically, or, often, into neighboring cells (FIGURE 3.16). Of these three methods, reuptake is by far the most common for most small neurotransmitters. Because of this rapid return of the concentration to normal levels, chemical neurotransmission is extremely precise.
Receptors W hen the neurotransmitter molecules are released into the synaptic cleft, they exert their effects by binding to recep- tors, specialized proteins in the membrane. Receptors can often be located presynaptically or on neighboring cells, but for now we’ ll concentrate on the over whelmingly most common ty pe: postsynaptic receptors. There are two main ways in which ty pical neurotransmitters transmit a signal to another cell: either by causing direct flow of ions into or out of the cell (ionotropic receptors) or by causing more indirect changes inside the cell by a cascade of signals (metabotropic receptors).
Let’s first look at the ionotropic receptors. There are dif- ferent concentrations of ions (charged particles) inside and outside the cells; thus, if you were to poke a hole in a mem- brane, ions would tend to flow in or out (their direction of flow depends on several factors that we will learn about in the next section). A n ionotropic receptor is essentially a sophisticated way of opening a temporary pore in the mem- brane. In its closed state, the receptor protein blocks the flow of ions; when it is opened, or gated, by the right type of neu- rotransmitter, the protein changes its shape and provides a pore in the membrane (FIGURE 3.15a). Many ionotropic recep- tors allow only a particular type of ion to pass through; thus a receptor that binds the neurotransmitter GA BA tends to selectively pass chloride ions, whereas a receptor that binds acetylcholine may selectively pass sodium ions.
The second type of receptor is called metabotropic and is also known as a “second-messenger-coupled” receptor (FIGURE 3.15b). To highlight some general principles, consider one well-studied family of such receptors, the G-coupled protein receptor. G-proteins are associated with the inside face of the postsynaptic membrane, and their function is to relay information from neurotransmitter receptors to other
NEUROTR ANSMIT TER CATEGORY E X AMPLE(S)
Monoamines Dopamine, epinephrine, norepinephrine, serotonin, melatonin
Amino acids Glutamate, aspartate, GABA, glycine
Peptide neurotransmitters Cholecystokinin, somatostatin, neuropeptide Y
Gases Nitric oxide, carbon monoxide
Organic cation Acetylcholine
TABLE 3.1.
Types of Neurotransmitters Ionotropic receptor
Extracellular fluid
Intracellular fluid
Neurotransmitter
Receptor
Effector
G protein
Receptor
Closed
Ion
Open
Metabotropic receptor
Neurotransmitter
(a) (b)
FIGURE 3.15 Two types of channels allow neurotransmitters to effect target cells. (a) Ionotropic receptors are opened—or gated—allowing ions to move through a passage in the membrane. (b) Metabotropic receptors relay signals to proteins inside the cell.
03-Eagleman_Chap03.indd 85 02/11/15 3:24 pm
86 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 86 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Postsynaptic Potentials Returning to the ionotropic receptors, what happens after neurotransmitters bind and cause ions to flow in or out of the postsynaptic cell? Because there are different concentra- tions of ions inside and outside the cell, there is a voltage dif- ference (also known as a potential difference) across the membrane. Normally, the outside of the cell is more positive than the inside, giving a resting membrane potential of about -70 millivolts (mV). Depending on the charge of the ions and in which direction they flow, the movement of ions across the membrane can make this potential difference smaller or larger. W hen positive ions, such as sodium, flow through a receptor into the cell (slightly reducing the difference between inside and outside), this is known as an excitatory postsynaptic potential, abbreviated as an EPSP (FIGURE 3.17a). Conversely, if neurotransmitter binding causes the potential difference in voltage between the inside and outside of the cell to grow larger (that is, the inside to become even more negative), this change in voltage is known as an inhibitory postsynaptic potential (IPSP) (FIGURE 3.17b). This can occur by allowing positively charged potassium to flow out of the cell or by allowing negatively charged chloride ions to flow into the cell.
Typical postsynaptic potentials are small changes in voltage—about 1 mV—and they are rapid, lasting only a few milliseconds. But recall that a typical cell in the cortex has about 10,000 synaptic inputs, meaning that the postsynaptic cell is receiving a symphony of signals at every moment. These small postsynaptic changes in the membrane poten- tial do not stay where they are, but instead funnel down toward the soma. Remember what you learned about the
(b) Diffusion
Synaptic vesicle
Synaptic cleft
Dendritic spine
Axon terminal
Voltage-gated Ca++ channel
Ca++
Neurotransmitter receptor
Post-synaptic density
Neurotransmitter
(a) Enzyme
degradation
(c) Neurotransmitter re-uptake pump
FIGURE 3.16 There are three ways by which neurotransmitters are cleared from the cleft. (a) Degradation, (b) diffusion, and (c) reuptake.
FIGURE 3.17 Postsynaptic potentials. (a) An excitatory postsynaptic potential (EPSP) occurs when positive ions flow through an ionotropic receptor into the cell, causing depolarization. (b) An inhibitory postsynaptic potential (IPSP) occurs when positive ions flow out of the cell, or negative ions flow in. This causes the difference in voltage between the inside and outside of the cell to grow larger, known as hyperpolarization.
(a)
Depolarization (EPSP)
Hyperpolarization (IPSP)
Time (ms) S
S Time (ms)
Cl–
K+
Extracellular fluid
Na+
Intracellular fluid
An EPSP is caused by Na+ entering the cell.
An IPSP can be due to K+ leaving the cell and/or Cl–
entering the cell.
–70
–73
0
V o
lt ag
e (m
V )
–65
–70
0
V o
lt ag
e (m
V )
(b)
03-Eagleman_Chap03.indd 86 02/11/15 3:24 pm
Synaptic Transmission: Chemical Signaling in the Brain 87
# 158305 Cust: OUP Au: Eagleman Pg. No. 87 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
animal, GA BA is considered an inhibitory neurotransmitter. Nothing changes about the molecule, only the consequences it has when it binds to its postsynaptic receptor.
For completeness, you should know one more thing about signaling between cells. A lthough chemical transmis- sion is the overwhelmingly common form of signal transmis- sion at synapses, another mechanism also exists: electrical synapses, also known as gap junctions, allow the direct pas- sage of an electrical signal from one cell to the next (Connors & Long, 2004). Such connections often allow the synchro- nized spiking of groups of neurons. Electrical synapses are far less common, and their function is less understood. For the rest of this chapter, we will concentrate only on chemical transmission, the typical mechanism of signaling at synapses.
soma: it summarizes the signals, and in the next section we will see how.
But first, one more important point about EPSPs and IPSPs: you will sometimes hear different neurotransmitters referred to as excitatory or inhibitory, meaning that they cause positive or negative changes in the membrane voltage. (For example, “glutamate is an excitatory neurotransmitter, whereas glycine is inhibitory”). But keep in mind that it is not the neurotransmitter molecule itself that is excitatory or inhibitory—it is the action of the receptor that determines the effect. For example, in the developing animal, GA BA is considered excitatory because its receptors cause EPSPs by allowing chloride ions to pass out of the cell. But as the animal grows older, the receptors allow chloride ions to pass into the cell, and GA BA now causes IPSPs. Thus, in the adult
THE BIGGER PICTURE: Psychoactive Drugs The release of neurotransmitters allows neurons to communicate with one another. Although it may not be obvious at first, this process of neu- rotransmission is the main target of drugs and medications that effect mood and cognition.
Drugs that affect the brain’s com- munication are known as psychoac- tive drugs, and these include drugs used to treat dementia and schizo- phrenia; drugs used to treat depres- sion or anxiety; and recreational drugs such as cocaine, marijuana, and even caffeine. Selective sero- tonin reuptake inhibitors (in common use as antidepressants) work by preventing serotonin from being transported back into the presynap- tic cell. As a result, they are classi- fied as agonists, because they cause more serotonin to be present in the cleft. Similarly, attention deficit hy- peractivity disorder medications prevent the reuptake of the neu- rotransmitters norepinephrine and dopamine. Some recreational drugs, such as cocaine, also act as agonists. In this case, cocaine works by pre- venting the reuptake of dopamine.
Another recreational drug, alcohol, works as an agonist to the neu- rotransmitter GABA and, as a result of stimulating GABA receptors, de- creases overall activity in your cen- tral nervous system.
Psychoactive drugs may also be antagonists, meaning that they dampen or block normal receptor function. In this class, we find some of the medications used to treat schizophrenia, which prevent dopa- mine from interacting with a certain type of receptor on the postsynaptic cell. One such medication, risperi- done, has a complex profile of ef- fects on many types of receptors. In addition to being a dopamine antag- onist, it is also an antagonist for cer- tain serotonin, norepinephrine, and histamine receptors. In terms of recreational drugs, caffeine and al- cohol are both antagonists. Caffeine works by blocking a neurotransmitter known as adenosine. Adenosine normally builds up during the day and, in these higher concentrations, causes drowsiness. Caffeine pre- vents adenosine from binding to its receptors and therefore helps you
stay awake. Alcohol, in addition to the agonist effects we mentioned above, is also an antagonist to the glutamate receptors, which results in memory impairments, among other side effects.
As you’ve seen here, psychoactive drugs can have many effects on your daily life. However, these descrip- tions only scratch the surface of a field better known as neuropharma- cology. As you saw with risperidone and with alcohol, a single drug can have different effects on different neurotransmitter systems. Further- more, a drug that is an agonist at one type of neurotransmitter receptor may also act as an antagonist at a different receptor. Because different subtypes of receptors are found in different parts of the nervous system, a given drug can end up en- hancing neural activity in one part of the brain, while at the same time damping down activity in other parts of the brain. For these reasons, trying to develop new drugs and un- derstand their effects is a tricky and time-consuming enterprise.
03-Eagleman_Chap03.indd 87 02/11/15 3:24 pm
88 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 88 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
larger voltage change, whereas an EPSP and an IPSP arriving at the same moment will cancel each other out (FIGURE 3.19). Because the soma receives hundreds or thou- sands of such signals at any moment, the total voltage of the cell is determined not by any one incoming signal, but in- stead by the overall pattern of all the inputs received all over the cell, both excitatory and inhibitory. If the number of
Spikes: Electrical Signaling in the Brain An electrode placed near a neuron reveals that from time to time the voltage across the neuron’s membrane suddenly reverses and then, about a millisecond later, is abruptly re- stored (FIGURE 3.18). This is known as an action potential, often called a nerve impulse or spike (R ieke, Warland, & Bialek, 1999). Spikes are all or none, meaning that they either happen or they do not, and they are always the same size. In this section, we will see how spikes happen, how they travel, and what they mean.
Adding Up the Signals The small voltage changes collected in the dendrites (EPSPs and IPSPs) travel along the dendritic membrane to the cell body, where all the branches come together. A lthough the postsynaptic potentials are small, they can add up with one another in two ways. First, signals that arrive at the soma at the same time (or even close to the same time) will add up when they reach the soma—this is known as temporal sum- mation. Second, signals that arrive on different branches of the dendrites will converge at the soma—this is known as spatial summation. As a result of both kinds of summation, the soma has the opportunity to integrate signals flowing into disparate parts of the dendrites.
Excitatory and inhibitory postsynaptic potentials add up like a simple math problem. Two EPSPs will sum to a
–70
–55
0
40
V o
lt ag
e (m
V )
Time (ms)
0 1 2
Stimulus
Threshold
Resting state
Depolarization Repolarization
Action potential
Failed initiations
Refractory period
3 4 5
FIGURE 3.18 A neuronal action potential, or spike, is a binary, all-or-none signal. It occurs when voltage across the neuron’s membrane suddenly reverses.
Subthreshold, no summation
Temporal summation Spatial summation Spatial summation of EPSP and IPSP
(a) (b) (c) (d)
Resting potential
Threshold of axon of postsynaptic neuron
E1E1E1E1
E1 E1 E1 E1+E2 E1+I1E1
E2
I1
–55 –70
0
M em
b ra
n e
p o
te n
ti al
( m
V )
FIGURE 3.19 Temporal and spatial summation. (a) No summation occurs when EPSPs arrive with a delay between them; they, individually, cannot drive the membrane voltage to the threshold for a spike. (b) Temporal summation occurs when EPSPs arrive close in time and their contributions add up at the soma, leading to an action potential. (c) Spatial summation occurs when signals arrive on different branches of the dendrites, converging at the soma. (d) If an EPSP and an IPSP arrive at different locations at the same time, they will cancel each other’s effect at the soma.
03-Eagleman_Chap03.indd 88 02/11/15 3:24 pm
Spikes: Electrical Signaling in the Brain 89
# 158305 Cust: OUP Au: Eagleman Pg. No. 89 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
how an Action Potential Travels Two ions play key roles in making an action potential: sodium (Na1) and potassium (K1). (In fact, there are several ions and proteins involved, but we can understand the big picture with only these two). W hen a cell is at rest, there is a high concentration of Na1 on the outside of the cell and a much lower concentration on the inside; this is exactly the opposite for K1 ions (FIGURE 3.21a).
excitatory potentials over whelms the number of inhibitory potentials, this can drive the voltage of cell toward more positive values, making it increasingly depolarized. If the cell voltage reaches a threshold, ty pically about -60 mV, something special happens: an action potential is generated at the axon hillock, the part of the axon that connects to the soma. The axon hillock is the most excitable part of the neuron and therefore the location where spikes are initiated (FIGURE 3.20).
(a) (b)
Dendrite
Axon
Axon hillock initial segment (action potential originates here)
Soma
Synaptic bouton
FIGURE 3.20 Because of its high excitability, the axon hillock serves as the spike initiation zone.
(a)
0
–55
30
–70
M em
b ra
n e
p o
te n
ti al
( m
V )
Time (ms)
(a)
(b) (c)
Action potential
Threshold
43210
Extracellular fluid
Intracellular fluid
Na+
K+
(b)
Na+
K+ K+
(c)
Na+
FIGURE 3.21 The sequence of a voltage spike. (a) At rest, there are more Na1 ions outside the cell than inside and more K1 ions inside the cell than outside. (b) When voltage-gated Na1 channels open, Na1 ions rush from the outside to the inside—both because of the concentration differences and because of the electrical field. (c) The depolarization caused by Na1 influx triggers the opening of K1 channels, which cause K1 ions to rush out, thus making the outside more positive again (repolarization).
03-Eagleman_Chap03.indd 89 02/11/15 3:24 pm
90 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 90 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
W hen the membrane potential rises beyond a certain threshold, it triggers the opening of voltage-gated ion chan- nels, in this case voltage-gated Na1 channels—ion channels that selectively pass Na1 and are opened only at particular voltages across the membrane. W hen these channels open, Na1 ions suddenly find a way into the cell. These ions are driven in by both the concentration gradient (there are many more on the outside than on the inside) and the elec- trical gradient (the inside of the cell is more negatively charged than the outside of the cell, attracting the positively charged Na1 ions into the cell) (FIGURE 3.21b).
Why doesn’t the axon become permanently depolarized and stay there? Because voltage-gated K1 channels are not far behind in their action. The influx of Na1 depolarizes the mem- brane further, which triggers the opening of the K1 channels (Kang, Huguenard, & Prince, 2000). Now, K1 ions flow down their concentration gradient (that is, there are more on the inside than on the outside, so they will tend to flow out) (FIGURE 3.21c). Because the K1 ions are positive and because they are rushing out of the cell, the inside becomes more negative—that is, it repolarizes. This return to a negative voltage shuts the voltage- gated Na1 channels and ends the swing in voltage.
This exchange of ions causes a voltage spike at the axon hill- ock, but how does an action potential travel? The answer is that the rapid voltage change gives just enough time and spreads far enough down the membrane for neighboring voltage-gated Na1 channels to open up, causing the same cycle of ion ex- change to happen nearby. In this way, the cycle of depolariza- tion and repolarization moves down the axon. By analogy, imagine mousetraps arranged in a long line. Each time one trap snaps shut, it gives just enough vibration to trigger the next one to snap and so on down the line. This way of passing the signal along the membrane is the trick by which action potentials propagate, and it is the main reason the nervous system can carry signals quickly across the distant expanses of the brain.
If you’re thinking ahead, you may ask why action potentials don’t travel in both directions, forward and backward. This is because there is a short refractory period after an action poten- tial, during which the Na1 channels are more resistant to open- ing. (In our analogy, it takes a bit of time to reset the mousetrap.) As a result, the action potential cannot move back to a location where it has already occurred, but can only travel forward.
For completeness we should mention that the full story of an action potential is slightly more complex, involving cal- cium and chloride ions. These ions contribute to the exact shape of action potentials, but they are not the major players. For our purposes, the level of detail in this section contains the principles you need to understand for the rest of the book.
Now that you understand how an action potential travels, you can understand what happened to the Japanese actor, Bandō Mitsugorō, whose story introduced this chapter. Tetrodotoxin molecules block the pore of the voltage-gated Na1 channel, thereby preventing the channel from opening and passing ions (Lee & Ruben, 2008). W hen there is no Na1 channel activity, there are no action potentials. Without action potentials, all communication stops. The fact that FIGURE 3.22 The nodes of Ranvier. (a) Diagram. (b) Microscopic view.
(a)
(b)
NR
NR
Node of Ranvier
Nodes of Ranvier
Layers of myelin
Axon
tetrodotoxin directly interferes with the most basic signaling mechanism of the nervous system—the action potential—ex- plains why it is 100 times more potent than potassium cya- nide, the poison of choice for most Hollywood screenwriters.
Myelinating Axons to Make the Action Potential Travel Faster Remember the myelination of axons that is done by the oli- godendrocytes (in the central nervous system) and the Schwann cells (peripheral nervous system)? Now that we’ve learned about action potentials, we are ready to turn to the function of the myelin sheath.
First, remember that myelination is not continuous along the length of the axon, but instead comes in segments of about 100 micrometers, giving the “string of sausages” look. The small gaps left between the myelin sheaths are known as the nodes of Ranvier, and it is at these points that ions from outside the cell can most easily flow in and out (FIGURE 3.22). In the stretches of myelin-insulated axon, it is difficult for ions to
03-Eagleman_Chap03.indd 90 02/11/15 3:24 pm
Spikes: Electrical Signaling in the Brain 91
# 158305 Cust: OUP Au: Eagleman Pg. No. 91 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
potentials. In addition to increasing the conduction veloc- ity of action potentials, myelination has the advantage of decreasing energ y expenditure: with fewer ions moving around, less energ y is needed to replace them (in other words, to “reset” the mousetraps). Most generally, myelina- tion is required for a vast network of neurons to function together properly, as seen by the stor y of champion speed skater Adam R iedy.
Action Potentials Reach the Terminals and Cause Neurotransmitter Release Action potentials travel down axons until they invade the axon terminals, and here our story comes full circle. The sudden voltage change in the terminal causes the opening of
move across the membrane, and the consequence of this is extraordinary: the action potential “leaps” directly from node to node instead of moving smoothly, as it does along an unmyelinated axon. This noncontinuous skipping of the spike is known as saltatory conduction (“saltatory” comes from the Latin saltare, “to jump”). The action potential is re- generated at each node, but not at the insulated stretches in between. The length of the myelination segments is just short enough that the depolarization at one node will be large enough to open the Na1 channels at the next node. To return to our previous analogy, this would be like placing mouse- traps far enough apart that the snapping of one gives just enough vibration to trigger the next—in this way, the signal reaches the end of the line much more quickly, skipping over a good deal of distance with each event.
The effect of saltator y conduction is to vastly increase the travel speed, or conduction velocity, of the action
CASE STUDY: Multiple Sclerosis In 2001, 20-year-old Adam Riedy was a strong candidate for the U.S. Olym- pic speed skating team. In the World Cup competition, he had just won a bronze medal in the 1,000-meter race and won two medals on relay teams.
But he awoke one morning and discovered his body didn’t feel quite right: “I found that the right side of my leg had gotten all tingly and numb. As the day went on, the pain moved all the way up my right side, then back down again” (Puet, 2002). He put down his concerns. The pain soon went away, allowing him to perform at full strength. But just before the Olympic trials in Utah, a more severe attack took place, weakening his right leg. Given his investment of 10 years of training, he tried to skate anyway—but it was to no avail. His body simply wouldn’t do what he wanted it to anymore.
Upon visiting a physician, Adam’s problems were diagnosed as multi- ple sclerosis (MS), an autoimmune
disease in which the immune system attacks healthy central nervous system tissue. It mistakes the body’s own healthy tissues as foreign tis- sues. The consequence of this self- attack is demyelination: the myelin sheaths become scarred (sclerotic) and are unable to perform their job of insulation. As a result, neurons of the central nervous system cannot prop- erly conduct signals, and the infor- mation flow is corrupted. Because the demyelination happens in small, isolated areas, patients often have varied and clinically isolated symp- toms. Complaints can include muscle weakness, difficulties with balance, problems with speech or vision, fa- tigue, and pain (Burks, Bigley, & Hill, 2009). As more areas of demyelin- ation accrue, the symptoms progress. Because the disease is progressive, it results in a gradual deterioration of neurological function.
Despite the ongoing research on MS, the cause of the disease re- mains unclear and there is no cure.
Strangely, the worldwide incidence of MS shows a clear pattern: the far- ther you live from the equator, the higher the chances of developing MS; this had led to the suspicion that the disease may have something to do with environmental factors such as climate, sunshine, and vitamin D (Disanto, Morahan, & Ramagopalan, 2012; Pierrot-Deseilligny & Souber- bielle, 2013). However, exceptions to this rule—such as the especially low incidence among the equatorially distant Maori of New Zealand or Sami people of Northern Europe— point to genetic involvement as well (Koch-Henriksen & Sorensen, 2010; Lin, Charlesworth, van der Mei, & Taylor, 2012).
Although MS is undergoing in- tensive study, one point remains clear: the integrity of the myelin sheaths is critical to the functioning of the nervous system. When myelin becomes damaged, the communi- cation networks in the brain become critically compromised.
03-Eagleman_Chap03.indd 91 02/11/15 3:24 pm
92 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 92 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
the synaptic cleft and interact with receptors on the next set of dendrites.
Neurons form dense networks. For simplicity, textbooks often draw a picture of neuron A transmitting a signal to neuron B. But in reality a single action potential traveling down an axon will invade each of its 10,000 terminals, caus- ing chemical release at a vast network of locations.
voltage-gated calcium channels, causing a rapid entry of cal- cium ions from the outside. The calcium ions cause the vesi- cles packed with neurotransmitter to fuse with the terminal membrane—and this causes the neurotransmitter molecules to spill out into the synaptic cleft (FIGURE 3.23). At this point we are back to the paradigm presented in the first section of this chapter: the neurotransmitter molecules now diffuse across
Action potential invades the axon terminal, opening Ca++ channels
The entry of Ca++ causes the vesicles to fuse with the membrane, allowing the neurotransmitters to be released
Action potential
Neuro- transmitter vesicle
Terminal
Postsynaptic cell
Voltage- gated Ca++
channel Ca++
NeurotransmitterReceptor
FIGURE 3.23 Action potentials lead to neurotransmitter release.
NEUROSCIENCE OF EVERYDAY LIFE: The Magic of a Local Anesthetic Why can the eye doctor put a solution in your eyes and then touch your eye- ball without it hurting? How can a den- tist extract a tooth while you’re awake? Why can a surgeon sew stitches into your leg while you watch? The answer involves the magic of a local anes- thetic (an 5 no, esthetic 5 feeling). Local anesthetics work simply by blocking action potentials. For exam- ple, lidocaine blocks the voltage-gated
Na1 channels, thus preventing action potentials from spreading along the axon. Touch and pain are signaled to your brain by action potentials travel- ing up the nerves of your peripheral nervous system to your central ner- vous system; they are then interpreted by your brain as a sensation (more on this in Chapter 6). When lidocaine is applied to a local area of the body, the electrical signals on the peripheral
nerves are blocked at their site of origin, so no sensations can arise. Al- though the receptors in your eyes, mouth, or skin are being activated, your brain has no way of getting the message. This simple but amazing technique highlights the fact that all of your body’s sensations are con- structed by the brain, a feat we will discuss in greater detail in later chap- ters of the book.
03-Eagleman_Chap03.indd 92 02/11/15 3:24 pm
What Do Spikes Mean? The Neural Code 93
# 158305 Cust: OUP Au: Eagleman Pg. No. 93 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
or selective, for par ticular stimuli. For example, imagine using an electrode to record the activ it y of a single neuron in the human brain (see Research Methods: Recording Action Potentials with Electrodes later in this chapter). You show photographs of several dif ferent t y pes of animals to the person—say a tiger, an eagle, and a rhinoceros. Let’s imagine that none of these pictures seems to “excite” the cell at all—that is, the cell generates no more action po- tentials beyond its normal background rate of activ it y. But now you present a picture of a mouse or a rabbit, and the neuron suddenly discharges many more action poten- tials (FIGURE 3.24; Mormann et al., 2011). In this case, the cell is described as being “selective” for par ticular stimuli over others. For example, specif ic regions of the primate brain contain neurons selective for faces (“ face cells”; Weiner & Grill-Spector, 2012); other regions contain neurons that respond v igorously to edges, colors, ani- mals, business logos, and hundreds of other stimuli (Bramao, Reis, Petersson, & Faisca, 2011; Kour tzi & Connor, 2011; Ungerleider & Bell, 2011). This is tr ue not only in v ision, but also more generally in hearing , touch, smell, v ibration, temperature, and so on: neurons are of ten tuned to par ticular stimuli and not to others (more on this in Chapters 5 and 6).
It is common to quantif y a neuron’s response in terms of its firing rate: the number of action potentials that occur per some unit of time (for example, “50 spikes per second ”). For example, in FIGURE 3.25, the response of a neuron in v isual cortex depends on the orientation of a bar. W hen the bar is rotated to a different angle, the indiv idual action potentials do not get bigger—rather, they become more frequent. That is, more of them occur in the same amount of time.
The idea that neurons encode stimuli by the number of action potentials in a small window of time is called rate coding. In this framework, neurons are specialized to detect certain stimuli, and the detection consists of a train of spikes. Because spike trains in a neuron can be elicited as a function
What Do Spikes Mean? The Neural Code If each spike occurring in your brain made a little popping noise, the result would be a cacophony that would blow your ears out. This is because most of your nearly 100 billion neu- rons are generating spikes somewhere between once per second up to a few hundred times per second. The sending of these electrical signals is the most energy-intensive process your brain engages in. It is costly, from an energy perspec- tive, because the ability to propagate action potentials re- quires the constant maintenance of chemical gradients (more sodium on the outside, more potassium on the inside). To return to our analogy, the system requires the “setting” of the long line of mousetraps. Once these traps have been snapped, the neuron needs to expend a good deal of energy to reset them. Specifically, it needs to push the Na1 ions back out and pull the K1 ions back in, a process that accounts for the majority of the energy expended by the brain (Ames, 2000). The fact that the brain is willing to spend so much of its resources on spikes suggests they carry important infor- mation. But what is that information? We will now explore how spikes—and trains of spikes—are thought to carry information.
Encoding Stimuli in Spikes W hat do spikes mean? It appears that spikes are the binary, all-or-none, fundamental letters of the nervous system, but what are these letters spelling?
It is common to measure the electrical activ it y of a single neuron in a human or research animal and watch the neuron’s electrical response to some stimulus in the outside world (say, show ing a v isual stimulus, or touching a f inger, or presenting a sound). Cer tain cells are tuned,
55 60 58 61 50 59
S p
ik es
p er
s ec
T ri
al n
u m
b er Spike
Time (sec)
FIGURE 3.24 A selective neuron responds with greater activity to one particular type of stimulus more than to other types. The blue dashes represent individual action potentials; different rows represent individual trials. The red histograms summarize the response of the neuron over many trials.
03-Eagleman_Chap03.indd 93 02/11/15 3:24 pm
94 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 94 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
is not a single spike, but the spike train (Ermentrout, Galan, & Urban, 2008).
A lthough such rate coding provides a good starting point for thinking about neurons, it is not the complete story, at least not every where in the brain. For example, ani- mals must react quick ly to threatening stimuli (think of how swiftly a fly zips away from a swat). If an animal always needs to count up spikes over some time window, that se- verely limits how quick ly it can decide what to do. Further, rate coding may be the wrong framework for encoding stim- uli that change quick ly, such as the position of predator or prey: by the time enough spikes have been averaged to tell a story, the message has changed. Reacting rapidly and en- coding changing stimuli are two tasks that describe much of what is important in an animal’s life. This suggests that the brain has methods of encoding stimuli that can go beyond rate codes.
Many possibilities ex ist for other coding methods. For example, although the rate coding hy pothesis assumes that the exact timing of indiv idual spikes doesn’t matter, studies on electric fish reveal that a response to a stimulus can have specific temporal structure (that is, the spikes are arranged precisely in time)—and that the temporal struc- ture could var y as a f unction of the stimulus properties, even while the total firing rate remains the same (Bullock et al., 2005) (FIGURE 3.26). That is, a neuron might respond w ith a dense burst of spikes to one stimulus and evenly spread spikes to another; alternatively, the timing of spikes in relation to the stimulus might shift earlier or later. In all these scenarios, the average firing rate remains the same. A s a result, rate coding would be too coarse a measure, fail- ing to capture the temporal properties of spik ing that carr y
of the specific properties of external stimuli, encoding is based on the average rate of firing over some time interval (Adrian, 1928).
W hy do neurons need to send a train of spikes? W hy can’t they simply send off a single spike at exactly the moment when it should occur? This may be for at least two reasons. First, physiologists observe trial-by-trial variability in the number of spikes and their timing when a sensory neuron responds to its preferred stimulus (Butts et al., 2010; Haslinger et al., 2012; Murphy & R ieke, 2006). In other words, the same stimulus results in a slightly different spike train each time. This variability supports the possibility that a neuron’s response to a stimulus on any given trial is essen- tially probabilistic: it is more likely to fire at certain moments than at other moments, giving the same general response but with some amount of variability from trial to trial (Knob- lauch & Palm, 2005). Second, most neurons spike occasion- ally even in the absence of a specific stimulus. For example, the neurons that carry visual information out of the retina send out several spikes per second even when it is completely dark (Korenbrot, 1995; Picones & Korenbrot, 1995). This spontaneous activity implies that a single spike is not a reli- able sign of the presence of a stimulus and perhaps, by itself, carries little information. Instead, neurons are “noisy”: whether a neuron fires at a specific moment or not is probabi- listic. In addition, the presence of a baseline firing rate allows the neuron to have a more flexible repertoire of responses, since it now has room to either increase or decrease its firing rate from the baseline. If the baseline were zero, the neuron would have no room to further decrease its activity in re- sponse to inhibitory stimulation. Given these consider- ations, it is generally thought that the relevant coding signal
Visual field
60
50
40
30
20
10
0
N eu
ra l r
es p
o n
se (
sp ik
es /s
ec )
Stimulus orientation (degrees) 0 45 90
FIGURE 3.25 The response of a cell depends on the orientation of the stimulus. A horizontal bar in the visual field inspires no action potentials, whereas a diagonal bar excites many action potentials. This cell can thus be said to be selective for diagonal bars.
03-Eagleman_Chap03.indd 94 02/11/15 3:24 pm
What Do Spikes Mean? The Neural Code 95
# 158305 Cust: OUP Au: Eagleman Pg. No. 95 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
information. There are several possibilities for coding beyond firing rate, and these may be used in different places and at different times in the ner vous system. We w ill explore some other methods involv ing populations of neu- rons shortly in this chapter.
Decoding Spikes W hen we consider the neural code—what the spikes mean—we must think from the point of v iew of a single neuron. Because most of the studies about spikes have fo- cused on outgoing spikes from single cells, it is easy to think that those spikes w ill carr y a good deal of informa- tion to the neurons they contact. But there’s a critical point to keep in mind about the nex t neurons contacted: a t y pical neuron in the cor tex receives some 10,000 sy nap- tic connections from other neurons. A single input is not suf f icient to generate a spike—instead, as we saw earlier, many input signals are needed w ithin a shor t inter val, and they must sum to create a suf f iciently large depolariza- tion. This suggests the possibilit y that a neuron receiv ing signals is not necessarily analy zing the details of each input line, but instead it is averaging over the population of its inputs. In this way, a neuron acts as a coincidence detector: it becomes activated by enough excitator y inputs coinciding in space and time to send it above threshold (FIGURE 3.27). In other words, an output spike is a response to the coincidence of many excitator y inputs ar- riv ing simultaneously (Abeles, 1982).
To envision this, imagine standing outside a sports sta- dium during a baseball game. You are unable to distinguish
Input: Natural stimulus
Electric organ discharge
Latency code
Burst duration code
Probability code
Microstructure code
Frequency code
Output: sensory nerve impulses
FIGURE 3.26 A palette of coding possibilities for carrying information about a stimulus (Bullock, 1968).
(a)
Presynaptic cell (sender)
Postsynaptic cell (receiver)
(b)
FIGURE 3.27 Are neurons integrators or coincidence detectors? (a) In the “assembly line” view of neurons, neurons pass messages to one another: the cell on the left is the sender, and the cell on the right integrates those signals as the receiver (Konig, Engel, & Singer, 1996). (b) Because neurons receive thousands of inputs, they may be better thought of as coincidence detectors. The cell body of the postsynaptic cell is unable to determine which presynaptic neuron sent which signal—instead, a postsynaptic spike will only signal the coincidence of many excitatory inputs arriving simultaneously.
the details of any of the indiv idual conversations of the 10,0 0 0 chatting fans inside. But when something special happens—like a home r un w ith bases loaded—all of the voices come into a sy nchronized unison, and you k now something impor tant has occurred. The sudden coinci- dence of the voices is how you can disting uish normal game play from a big moment, and it is presumably the way that neurons operate. Neurons are not driven by other, single neurons, but instead by activ it y patterns over a population. We are now ready to turn to populations of neurons.
03-Eagleman_Chap03.indd 95 02/11/15 3:24 pm
96 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 96 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
RESEARCH METHODS: Recording Action Potentials with Electrodes Action potentials are typically mea- sured in a branch of electrophysiol- ogy known as single-unit recording (FIGURE 3.28). In this technique, a fine, thin piece of wire (usually made from tungsten or a platinum–iridium alloy) is inserted into an animal’s brain. Because the tip of the elec- trode rests near one or more neu- rons, the action potential in the cell will displace ions and generate a signal in the electrode. With this technique, one can detect each time a spike occurs in a nearby neuron. It was by this technique that research- ers Torsten Wiesel and David Hubel recorded activity of single neurons in cats and discovered the principles of organization of the visual cortex (we’ll get to this in Chapter 5) (Hubel & Wiesel, 1962; Roy & Wang, 2012). For this work, they won the Nobel Prize in 1981.
Although this technique is usu- ally confined to animal research, it is sometimes feasible to implement in humans who are undergoing neu- rosurgery. In a study by Itzhak Fried and colleagues, an electrode was inserted into the hippocampus of an epilepsy patient (Gelbard-Sagiv, Mukamel, Harel, Malach, & Fried, 2008). As you learned in Chapter 2, the hippocampus is involved in memory. The patient was then shown several short video clips—for exam- ple, part of a Martin Luther King speech, a clip from The Simpsons television series, a flyover of the Hol- lywood sign, and more. The hippo- campal neuron responded vigorously to certain video clips—in this case,
the Simpsons clip—but not to others. Remarkably, when the patient was later asked to recall what he had seen, the neuron again fired vigorously when the patient thought about the Simpsons clip.
Let’s think about what this find- ing means. First, it does not indicate that this neuron is the neuron that encodes The Simpsons—instead, it indicates that the neuron is part of a larger network that is activated by The Simpsons (either viewing or re- calling). Second, it does not mean that the neuron belongs only to the “Simpsons network” of neurons: instead, neurons can participate in different coalitions at different times, as we will see shortly. At one moment, this neuron may be in- volved with thousands of other neu- rons in recalling The Simpsons, and the next moment it may be involved in another vast pattern of activity that represents the sight of a cell
phone resting on a chair. In fact, the neuron that responded to the Simpsons clip was also found later to respond to a clip of the Seinfeld television series.
When thinking about single-unit recording, it is critical to keep in mind that the activity of single neu- rons merely represents a piece of a much larger pattern. To overcome the limitations in interpretation of single neurons, more recent approaches have implemented mul- tielectrode recording, in which a collection of thin electrodes are bundled together to record the ac- tivity of up to hundreds of neurons at once (Whitson, Kubota, Shimono, Jia, & Taketani, 2006). Although this is a move in the right direction, some researchers still lament that there is still no technology at the “sweet spot” level of measuring the detailed spiking activity of tens of thousands of cells simultaneously.
Microelectrode
Neuron
FIGURE 3.28 A microelectrode for recording from a single neuron. The tip is usually just a few micrometers wide.
03-Eagleman_Chap03.indd 96 02/11/15 3:24 pm
Individuals and Populations 97
# 158305 Cust: OUP Au: Eagleman Pg. No. 97 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
other lines of evidence, the graceful degradation of memory speaks against a grandmother cell strategy for the brain.
But what is the alternative?
Populations of Neurons A new understanding emerged when researchers began to see that systems of simple units—when str ung together properly—could display remarkable emergent properties. For example, the representation of an object, such as a face, can be distributed across neurons in a population, and that distrib- uted representation is both a realistic and a flexible strategy for the nervous system (Lehky & Sereno, 2011; Pasupathy & Connor, 2002; Reddy & Kanwisher, 2006). In the framework of population coding, recognition of something—say, your neighbor or college professor—is achieved by a coalition of neurons: a group of some hundreds or thousands of neurons temporarily working together as a team (Lehky & Sereno, 2011; Pasupathy & Connor, 2002; Reddy & Kanwisher, 2006). A given neuron might participate in many different coalitions depending on the task and the occasion. As an analogy, con- sider that a person can at one moment attend a rally for the Democratic Party, at a later moment attend a volunteer group meeting for building homeless shelters, and later on serve on an organizing committee for a fundraiser. The same person is contributing to different coalitions at different times.
Similarly, thinking from the point of view of popula- tions, each neuron contributes its piece to larger patterns at any given moment (deCharms, 1998). If we found a neuron in your brain that reliably responds when you see your grandmother, that neuron would not correctly be described as a grandmother cell—instead, its response would be un- derstood as its standard contribution to the population rep- resenting your grandmother. That same neuron will also make contributions (perhaps at different times and with dif- ferent spike patterns) to the recognition of other, different stimuli, such as nail clippers, balloons, or ducks.
So what does an activity level in an individual neuron mean? In the population view, the neuron’s response is simply the piece that the neuron contributes to the pattern of the pop- ulation as a whole (Lehky & Sereno, 2011). In the case of face representation, for example, a single neuron may not represent anything identifiable (such as a nose or an eye)—it may, rather, be playing a more diffuse and nonintuitive role in the represen- tation. That is, we should not expect that we can describe the individual neuron’s representational capacities using the same vocabulary we use to talk about what is perceived in everyday life, such as noses, eyes, mouths, and so on.
Population coding also permits a vastly greater range of representations than could be achieved if each neuron coded for a single property. For a simple illustration, imag- ine that you had 10 neurons, each of which can have three different levels of activity, or firing rates: low, medium, and high. With this simple system, 1,000 (103) different stimuli can be encoded, compared to just 10 stimuli if activity in
Individuals and Populations The experimental success in discovering selectivity in indi- vidual neurons originally led to a view of the nervous system as a confederation of specialist neurons, in which each neuron encodes some feature of the world. This idea, known as local coding, postulates that all stimuli in the outside world become represented uniquely by different neurons (Churchland & Sejnowski, 1992). For example, activity in a given neuron may represent a particular geometrical shape, whereas another neuron might represent a particular animal and another still might represent a more complex stimulus, such as NFL quar terback Ben Roethlisberger throw ing a football.
At least some brain areas contain neurons that are highly selective in their representations of sensory stimuli. For ex- ample, studies in neurosurgical patients have found individual neurons in the medial temporal lobes that respond specifically to the name or photograph of particular actors, family mem- bers, or famous landmarks (Quiroga, Kreiman, Koch, & Fried, 2008). These observations have raised the question of whether the brain might contain a single unique “grandmother cell” for your grandmother and presumably a single unique cell for every other familiar individual in your life experience: a local code of one unique cell to one unique stimulus.
But several difficulties appear to be fatal to the idea of local coding. First, although there are a great number of neu- rons in the brain, there are almost certainly not enough to recognize all the distinct patterns a person can recognize in a lifetime. Consider all the highly distinctive fonts in which you can read the word “dog.” Does the nervous system need a distinct neuron for each one? And consider the comparable issue for the many faces in many orientations in many light- ing conditions that are easily recognizable by a single brain over a lifetime. The arithmetic suggests that nervous systems must have evolved a more powerful and flexible strategy than local coding, although it might be used sparingly for special purposes. Consider also how lucky the neurosurgeons would had to have been to find the one unique grandmother cell among the hundreds of millions of neurons in each pa- tient’s medial temporal lobe: a task akin to finding a needle in an entire field of haystacks (Waydo, Kraskov, Quian Quiroga, Fried, & Koch, 2006).
Another major problem with local coding is the fact that brain cells die naturally throughout your lifetime, which would leave local-coding memory vulnerable to damage and degradation. If specific individuals were encoded in single cells, you might expect that as you got older and these cells were lost, specific individuals would abruptly vanish one by one from your memory, like photographs dropping out of an old photo album. Yet this is not what happens—instead, memory degrades steadily and “gracefully” over time during the progress of senile dementia (A lmor et al., 2009). As with
03-Eagleman_Chap03.indd 97 02/11/15 3:24 pm
98 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 98 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
A second possibility for forming a coalition is to use the time domain: in this framework, neurons that are members of a temporary coalition fire synchronously, thereby distin- guishing them from other neurons that are active for differ- ent reasons (Amari, Nakahara, Wu, & Sakai, 2003). In this theoretical framework, those neurons representing a cup of coffee—its smell, its taste, its sight—will temporarily coop- erate with each other as a coalition by firing spikes at the same time. A s before, a neuron can par ticipate in entirely dif ferent coalitions from moment to moment by the sy n- chrony of its spikes with other neurons’ spikes.
In one of the first experimental examples of synchrony, investigators found that nearby neurons that encode similar properties (say, the orientation of a bar) tend to fire spikes syn- chronously (Gray & Singer, 1989). That raised the possibility that neurons at even more distant locations could form tem- porary coalitions to encode features, simply by firing synchro- nously. In one demonstration of this principle, researchers found that distant neurons in the visual cortex fire more syn- chronously when a single bar is presented than a bar broken into two halves with a gap between them— presumably en- coding that the first condition represents one single object (Engel, Fries, & Singer, 2001; Engel & Singer, 2001). Note that if a physiologist were averaging spikes over hundreds of milliseconds or so, the subtle details of synchronous firing would be missed—in other words, a typical rate-coding anal- ysis of the data would be blind to the existing synchrony.
Another example of synchrony can be demonstrated in the auditory system. When a tone comes on and stays on, neurons in the auditory cortex give a brief response at the beginning and sometimes a brief response at the offset (FIGURE 3.30a) (Gillespie & Walker, 2001). This left open an important question: how does the brain encode that the tone has come on and remains on? The answer turned out to be that although there was no change in firing rate during the continuous presence of the tone, simultaneously recorded neurons change their relative timing, becoming more synchronized with each other while the tone is on (deCharms & Merzenich, 1996) (FIGURE 3.30b).
Again, the fact that these cor tical sites change their relative timing without changing their average firing rate is
each neuron simply “stood for” a single stimulus. Popula- tion coding is thus much more versatile than local coding (Churchland & Sejnowsk i, 1992; Lehky & Sereno, 2011). As we will see in Chapter 5, human color vision is a stun- ning example of how the brain can get by with only three ty pes of color detectors in the eye, but we can nevertheless distinguish upward of 10,000 hues. This is because the rep- resentation of color depends on the relative activity of the three ty pes of detectors distributed across the population.
Finally, population coding allows the nervous system to average over the problem of noise in individual neurons. Pre- cision appears to be achieved on any one trial by the activity of many similar neurons. That is, noisy neurons use popula- tion coding to achieve precision, and this strategy is more ef- ficient than the use of nonnoisy precise components (Deneve, Latham, & Pouget, 2001). Future computers may shift in the biological direction using a higher number of noisy compo- nents rather than a smaller number of precise components.
Forming a Coalition: What Constitutes a Group? We have seen the advantages of neurons cooperating in a dis- tributed coalition. But how exactly do the neurons form into teams? There are at least two ways this can happen. First, neu- rons can become active in such a way that each neuron mutually excites the others. The simplest version consists of two neurons releasing excitatory neurotransmitters on one another; scale this up to imagine networks of tens or hundreds or thousands of neurons temporarily maintaining high firing rates because of their mutual connectivity (Koch & Crick, 2001) (FIGURE 3.29). The central idea is that a population that supports one another’s high firing rate is a temporary coalition. As you can see, the term “coalition” here parallels the way it is used in political par- ties, when groups of people support each other for a common cause. Note that in the next moment, any individual neuron may participate in a completely different coalition, contributing to supporting a high firing rate in a new group.
FIGURE 3.29 Neurons that excite each other can form coalitions. (a) Two neurons that mutually excite one another. (b) A larger coalition of excitatory neurons.
(a)
(b)
Excitatory connections
03-Eagleman_Chap03.indd 98 02/11/15 3:24 pm
Individuals and Populations 99
# 158305 Cust: OUP Au: Eagleman Pg. No. 99 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
“noise.” However, we suspect neuroscience may someday be able to interpret this ubiquitous activity with more theoretical precision. For example, consider that the brain never starts a task from complete rest, but instead is always active. The level of activity existing when a task begins may be better under- stood as the context upon which the incoming data arrives, and this context may include information about recent experi- ence, past rewards and punishments, and associated internal and external variables (including feeling ill, well, energetic, bored, irritable, anxious, jumpy, excited, and so on).
Another issue to consider is that current neuroscience con- centrates on the details of individual neurons and spikes, as we have in this chapter. But a wider view of “neural activity” includes several other types of activity beyond spikes. For example, there are nonspiking neurons found in such places as the retina and hypothalamus. These generate electrical signals that can have different amplitudes (rather than all-or-none spikes), and they certainly appear to be involved in coding stimuli (Victor, 1999). Next, it remains an open question whether, in addition to neu- rons, some glial cells have an information-processing role, from allowing neurotransmitters to stay longer in the cleft to encour- aging the formation of new synapses (Araque, 2008). Finally, complex biochemical cascades are constantly churning inside of neurons. Although our technologies to examine these interac- tions are limited, we already know that these intracellular net- works are as complex as neural networks—just much smaller.
Next, much is still unknown about spike encoding and de- coding, and in part this is because of the historical details of how the problem has been studied. Early ideas about neuronal coding owed much to the details of the canonical experimental setup. First, recording is done from only a handful of cells at a time, and second, a simple, nonbiological stimulus is typically used. A typical stimulus for studying hearing might involve a simple beep; for the visual system, it might be a diagonal bar of light moving from left to right. A natural scene, such as the sounds and sights of forest life, was (and still is) considered filled with confusing variables and has been traditionally avoided (Eggermont & Ponton, 2002; Super & Roelfsema, 2005).
Another feature of traditional experiments, standard even through the 1980s and still around today, was to anesthetize the animal lightly but sufficiently to keep it still, thereby permitting an accurate recording. Although the anesthesia did reduce movement, it was itself a kind of confound, since anesthetic agents change the way cells interact and respond (Alkire, Hudetz, & Tononi, 2008; Andrada, Livingston, Lee, & A ntognini, 2012). Moreover, it was problematic that al- though particular cells displayed a regular response to the stimulus, the anesthetized animal presumably was not really perceiving it (Alkire et al., 2008; Andrada et al., 2012). An additional problem was that in the absence of a behavioral response to the stimulus, it was hard to tell whether the stim- ulus was significant for the animal’s brain as a whole.
Current technologies give us access to spikes from small populations of cells, but much of neural coding might be hap- pening elsewhere—for example, as temporal codes between vast assemblies of neurons throughout the brain. The possi- bility that information might be encoded in the temporal
something that would have been missed if we thought only about rate coding by single neurons.
Further experiments concerning synchrony across cor- tical areas have yielded mixed support for this framework, so the full story of synchronous firing is not a closed case (DeWeese & Zador, 2006; Samengo & Montemurro, 2010). But the synchrony framework offers one plausible possibility for how distant groups of neurons form a coalition.
Open Questions for Future Investigation Even after the 1990s were declared “The Decade of the Brain,” you are entering a field in which there are currently more questions than answers. The neural code—that is, the mean- ing of spikes in the brain—remains unsolved. One issue of confusion for neuroscientists is that neurons seem to exhibit a great deal of seemingly random activity, including membrane voltage fluctuations and spontaneous spikes. The classical paradigm deals with this activity mainly by assuming it is
FIGURE 3.30 Changes in the mean firing rate and neuronal correlation in the primary auditory cortex of a monkey. (a) A continuous tone (top) was presented while recording from two sites in the auditory cortex. At both sites, the average firing rate showed only a little rise in activity at the onset and offset of the sound. (b) The synchrony of the spikes across the two sites (during an initial silent period, blue, and during the sound, red) reveals that the spikes in the two areas become more synchronized with one another while the sound is present.
(a)
(b)
300
200 100
0
100
50
0
S p
ik es
p er
s ec
o n
d S
yn ch
ro n
iz at
io n
b et
w ee
n n
eu ro
n s
During soundDuring silence
Electrode 1
Sound on
Electrode 2
03-Eagleman_Chap03.indd 99 02/11/15 3:24 pm
100 PART 1 • ChAPTER 3 Neurons and Synapses
# 158305 Cust: OUP Au: Eagleman Pg. No. 100 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Tetrodotoxin blocks action potentials by blocking the pore opening of the fast, voltage-gated Na1 channels. Although the soma tries to generate an action potential, the axon cannot transmit the spike because the Na1 channels cannot pass ions: they have been rendered useless, and the nervous system falls silent. Besides complete paralysis of the skeletal muscles, the diaphragm muscles cannot move, and suffocation ensues. The example of the Japanese Kabuki actor underscores that the molecular details matter: the brain sends signals out from its command center in the skull to the entire rest of the body, and without proper conduction of electrical signals and proper neurotransmission, the system does not operate.
In upcoming chapters we will see how various drugs of abuse interact at all stages of the neural signaling pathways, from neurotransmitter release, postsynaptic receptor func- tion, and action potential propagation. For now, as you come to appreciate the function of neurons, consider how many places along the chain there are for the system to become modified by chemicals.
relationships between neurons is likely, but a better under- standing will require new techniques that involve measuring from hundreds, thousands, or eventually millions of neurons at once. This will allow us to finally understand how brain tissue acts as a new kind of physical medium through which signals move. The idea is that as signals spread through this medium, computations are performed by the interaction of patterns riding upon the medium. In other words, the indi- vidual neurons don’t provide the neural code; they provide the surface on which the neural code rides.
Conclusion We started this chapter with the story of Bandō Mitsugorō, the Japanese Kabuki actor who daringly (and not so wisely) ate the forbidden livers of puffer fish and died. With your new k nowledge, you can now understand what happened.
KEY PRINCIPLES
• Neurons and glial cells are the basic building blocks of the nervous system.
• Neurons consist of four important zones: (1) den- drites, which receive the chemical signals and chan- nel them to the soma, (2) the soma, which summarizes the incoming signals, (3) the single axon, which con- veys action potentials away from the soma, and (4) axon terminals, which transmit chemical signals.
• Neurons are linked together in dense networks, connected to each other by synapses, the sites of chemical transmission.
• Glial cells help with breaking down neurotrans- mitters, wrapping myelin sheaths around axons to speed electrical signaling, and regulating the chemical environment around neurons.
• Neurotransmitters diffuse across the synaptic cleft and bind to receptors on the postsynaptic target.
• Dendrites decode information by responding with small graded voltage changes to neurotransmitter behavior on the membrane; the dendrites and soma sum these signals. Output depends on whether the summed voltage reaches the thresh- old for initiating a spike.
• A neuron codes information not with single spikes, but instead in its frequency of firing (rate coding).
• Individually, neurons are noisy; collectively, they can be precise.
• Although neurons are traditionally recorded from one at a time, neural coding involves populations of neurons working together in transient coalitions.
KEY TERMS
The Cells of the Brain neuron (p. 77) membrane (p. 77) dendrites (p. 77) soma (cell body) (p. 78) nucleus (p. 78) axon (p. 78)
axon terminals (p. 78) afferent neuron (p. 79) efferent neuron (p. 79) multipolar neurons (p. 80) bipolar neurons (p. 80) monopolar neurons (p. 80) glial cells (glia) (p. 80)
oligodendrocytes (p. 80) Golgi staining (p. 81) Nissl staining (p. 81) autoradiography (p. 81) immunocytochemistry (p. 81) DNA (p. 81) in situ hybridization (p. 81)
03-Eagleman_Chap03.indd 100 02/11/15 3:24 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 101 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical-Thinking Questions 101
REVIEW QUESTIONS
1. How did the discovery that brain cells are dis- crete lead to new ideas about neural signaling?
2. Action potentials are all or none. Does that mean that brains are like digital computers? Why or why not? In what way are signals be- tween neurons not like digital computers?
3. What is the advantage of myelin for an axon?
4. What is the sense in which neurons “make decisions”?
5. Why is it correct to think of a neuron as a pattern recognizer?
6. What are the advantages of having populations of neurons encode stimuli rather than individual “grandmother cells”?
7. What are two possible ways in which neurons might combine to form a coalition?
Schwann cells (p. 82) myelin (p. 82) myelin sheaths (p. 82) nodes of Ranvier (p. 82) astrocyte (p. 82) microglia (p. 82)
Synaptic Transmission: Chemical Signaling in the Brain
synaptic cleft (p. 83) synaptic vesicles (p. 83) acetylcholine (p. 84) monoamines (p. 84) catecholamines (p. 84) amino acids (p. 84) glutamate (p. 84) GABA (gamma-aminobutyric
acid) (p. 84) peptide neurotransmitters (p. 84) retrograde transmitters
(p. 84) receptors (p. 85) ions (p. 85)
ionotropic receptors (p. 85) metabotropic receptors (p. 85) G-coupled protein receptor
(p. 85) G-proteins (p. 85) second messengers (p. 85) ion channels (p. 85) degradation (p. 85) reuptake (p. 85) transporters (p. 85) membrane potential (p. 86) excitatory postsynaptic
potential (EPSP) (p. 86) inhibitory postsynaptic
potential (IPSP) (p. 86) electrical synapses (gap
junctions) (p. 87) agonists (p. 87) antagonists (p. 87)
Spikes: Electrical Signaling in the Brain
action potential (nerve impulse or spike) (p. 88)
temporal summation (p. 88) spatial summation (p. 88) depolarized (p. 89) threshold (p. 89) axon hillock (p. 89) voltage-gated ion channels (p. 90) concentration gradient (p. 90) electrical gradient (p. 90) refractory period (p. 90) saltatory conduction (p. 91)
What Do Spikes Mean? The Neural Code
rate coding (p. 93)
Individuals and Populations local coding (p. 97) population coding (p. 97) coalition (p. 97)
CRITICAL-THINKING QUESTIONS
1. Given everything that you have learned about the nervous system and the advantages of myelina- tion, why do you think that many neurons are un- myelinated? Do you think that myelination serves a more critical role in some parts of the nervous system than in others? Explain your answer, using examples to illustrate your reasoning.
2. From an evolutionary perspective, do you think it would be better to have cells that functioned
more like “grandmother cells” to recognize ob- jects or cells that worked together in coalitions? Explain your reasoning.
3. In what other ways, besides the ones described in this chapter, do you think that neurons could com- bine to form a coalition? In general terms, de- scribe how you would investigate whether neurons combined to form a coalition in these ways.
03-Eagleman_Chap03.indd 101 02/11/15 3:24 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 102 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
102
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Explain how the brain can reconfigure its circuitry
to adapt to changes in sensory input.
• Illustrate how the brain can reconfigure its circuitry to enable new forms of behavior and how neuromodulators control plasticity according to information relevance.
• Describe how the brain adapts to losing neural tissue.
• Explain why the brain has a sensitive period for learning.
• Characterize how genetic factors and experiences interact in brain development.
• Illustrate two biological mechanisms of neural competition.
• Characterize a rapid and a slow mechanism for changing neural circuitry.
• Show how the brain incorporates new forms of sensory input.
04-Eagleman_Chap04.indd 102 02/11/15 3:27 pm
103
# 158305 Cust: OUP Au: Eagleman Pg. No. 103 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Neuroplasticity STARTING OUT: The Child with Half a Brain
The Brain Dynamically Reorganizes to Match Its Inputs
CASE STUDY: Phantom Sensation
RESEARCH METHODS: Mapping Out the Brain
The Brain Distributes Resources Based on Relevance
NEUROSCIENCE OF EVERYDAY LIFE: Pianists and Violinists Have Different Brains
CASE STUDY: The Government Worker with the Missing Brain
The Brain Uses the Available Tissue A Sensitive Period for Plastic Changes
CASE STUDY: Danielle, the Feral Child in the Window
Hardwiring versus World Experience The Mechanisms of Reorganization Changing the Input Channels
CASE STUDY: The Man Who Climbs with His Tongue
THE BIGGER PICTURE: Adding New Peripherals
CHAPTER 4
04-Eagleman_Chap04.indd 103 02/11/15 3:27 pm
104 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 104 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT: The Child with half a Brain By the time Matthew S. was 6 years old, he began to have several epi- leptic seizures each day—some- times every few minutes. Medications were of no use. He was diagnosed with Rasmussen’s en- cephalitis, a rare, chronic inflam- matory disease that typically affects only a single brain hemisphere. His parents explored their options and were shocked to learn that there was only one known treatment for Rasmussen’s: removal of an entire hemisphere of the brain (Borgstein & Gottendorst, 2002).
But how could it be possible to live with half of the brain missing? Aren’t the functions of the brain dis- tributed widely across its territo- ries? Wouldn’t removal of one half be fatal—or at least devastating to Matthew’s quality of life?
With no remaining options, Mat- thew’s parents took him to Johns Hopkins Hospital in Baltimore, Maryland, where he underwent a hemispherectomy: the complete removal of half the cerebrum (FIGURE 4.1). The empty half of the skull filled up with cerebrospinal fluid, which shows up as a black void in neuroimaging.
Matthew walks with a slight limp on the opposite side of his body. Otherwise, he lives a normal life with almost no measurable deficit in cognition or behavior. How
can this be possible? Because the remainder of his brain has dynami- cally rewired to take over the miss- ing functions. The normal maps of the brain have redrawn themselves
on a smaller piece of neural real estate. How the brain accomplishes this remarkable feat—something no manmade machine can yet do— is the subject of this chapter.
FIGURE 4.1 Hemispherectomy. In a hemispherectomy, half the brain is surgically removed. This surgery has become standard operating procedure for Rasmussen’s encephalitis, a rare inflammatory disease that often affects only one hemisphere. Amazingly, as long as the surgery is performed before the age of 8, the child does remarkably well: the remainder of the brain dynamically rewires to take over the missing functions.
The brain is often thought of as a fixed organ with different regions dedicated to specific tasks. But the brain is better un- derstood as a dynamic system, constantly modifying its own circuitry to match the demands of the environment and the goals of the animal. This ongoing rewiring is the brain’s most fundamental principle and the source of its utility. W hereas your computer is built with hardwiring that remains fixed
from the assembly line onward, the brain dynamically recon- figures, ever so subtly, with each new experience. It reorga- nizes itself from the level of molecules in the synapses to the level of the gross anatomy visible to the naked eye. W hen you learn something new (such as your professor’s name), your brain physically changes. This ability to physically change, and to hold that change, is known as plasticity—just like the
04-Eagleman_Chap04.indd 104 02/11/15 3:27 pm
The Brain Dynamically Reorganizes to Match Its Inputs 105
# 158305 Cust: OUP Au: Eagleman Pg. No. 105 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The Brain Dynamically Reorganizes to Match Its Inputs In the sensory and motor areas of the cortex (which we will dis- cuss in the next chapters), neighboring populations of neurons generally represent neighboring parts of the body—that is, the hand is represented near the forearm, which is represented near the elbow, and the upper arm, and so on (Nakamura et al., 1998). This map of the body is known as the homunculus, or “little man” (FIGURE 4.2). But how does your brain, encased in its dark vault of silence, know what your body looks like? Is the body plan genetically prespecified? And what would happen if your body changed—say, by the loss of an arm or the addition of two more legs: would the homunculus change?
Changes to the Body Plan Changes to a body plan—such as the loss of a limb—lead to massive cortical reorganization. In one study with monkeys, the nerves carrying signals from one arm were severed— this is known as deafferentation. W hen the monkeys were tested 12 years later, it became clear that their somatosen- sory cortex had rearranged: the areas once representing the
material we call plastic, which can be molded and retain its new shape. Plasticity is the basis of learning and memory, as we’ll see in Chapter 9. In this chapter, we will discover how the principles of plasticity allow the brain tremendous flexibility: using the strategy of reconfiguration by experience, a brain can find itself within any body plan (four arms, eight legs, and so forth), and it will figure out how to configure itself to con- trol it optimally. A brain can find itself in any ecosystem (e.g., jungle, swamp, or mountains), and it will learn how to move in it. A brain can find itself in any country, and it will absorb the local language and culture. In this chapter, we’ll find out how.
First, we’ll examine several examples that unmask the plastic capabilities of the brain by examining how it responds when the sensory input changes. Next, we’ll discover how plastic changes are tied into relevant goals for an animal, and we’ll see how the brain distributes its functions according to the available territory. This sort of plasticity happens mostly within a time window known as the sensitive period. We will then be ready to confront the nature-versus-nurture question to learn how much of the brain is prewired by genet- ics and how much is plastic. Finally, we’ll turn to the me- chanics of the reorganization to understand what is happening at the level of the synapses and neurons. The ex- amples we will find along the way are all captured by a single organizing principle: the brain distributes its resources ac- cording to what’s important for the organism, and it does so by having do-or-die competition at every level, from neurons to brain regions.
FIGURE 4.2 (a) Motor homunculus and (b) sensory homunculus. The body becomes topographically mapped on the precentral gyrus (motor cortex) and postcentral gyrus (somatosensory cortex). Those areas with more sensation, or that are more finely controlled, have larger areas of representation.
Motor cortex Somatosensory cortex
(b) Sensory homunculus
(a) Motor homunculus
Intra- abdominal
Swallowing Tongue
Jaw
Lips
Face
Eyelid and eyeball
Brow
Neck
Thumb
Index
M iddle
Ring Little
H and
W rist
Elbow Shoulder
T run
k
H ip
K n
ee A
n kle
T o
es
Pharynx
Teeth, gums ,
and jaw Tongue
Lips Lower l
ip
Uppe r lipFa
ce Nos
eE ye
Th um
b In
de xM
id dl
e
Ri ngL
itt le
H an
d
W ris
tEl bo
w Fo
re ar
m
A rmSh
ou ld
er
H ea
d
N ec
k
T ru
n k
H ip
L egT o
es Fo
ot
G en
it al
s
04-Eagleman_Chap04.indd 105 02/11/15 3:27 pm
106 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 106 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
CASE STUDY: Phantom Sensation D.M., a 43-year-old woman, had her right arm amputated just above her elbow. Although her arm was miss- ing, she insisted that she could still feel her right forearm, hand, and fin- gers. She could, in fact, “move” them. When she walked, she felt her
missing hand swinging correctly with the motion of her body. She reported that the sensation from her phantom arm felt a bit colder; otherwise, it was mostly indistinguishable from the sensation in her existing arm. When the right side of D.M.’s face
was touched, she felt as though her missing limb was being touched at the same time. As the year after her amputation progressed, she felt as though her hand and fingers moved closer to the stump of her arm (Hal- ligan, Marshall, & Wade, 1994).
FIGURE 4.3 Changes in sensory maps: the brain adapts to changes in incoming activity, even in adulthood. After hand amputation in humans, neighboring cortical territory (purple and green) takes over the territory that previously coded for the hand (orange).
Upper arm/trunk/leg
Normal somatosensory cortex Amputee somatosensory cortex
Hand/forearm Face
FIGURE 4.4 Phantom sensations in amputee. When an arm is amputated, a touch to the face (represented in neighboring cortical territory) often results in a sensation that the phantom arm is being touched. This 17-year-old man lost his right arm. When a cotton swab was stroked on his face one month later, the area labeled “T” gave rise to sensation in his phantom thumb, “P” to his phantom pinkie, “I” to his index finger, and “B” to the ball of his thumb. Figure from Ramachandran and hirstein (1998).
arm had been taken over by the neighboring areas, which happened to respond to touch on the face (Pons et al., 1991).
This same phenomenon is observed in humans. When an arm is amputated—say, after a motorcycle accident—the neighboring face representation creeps into the neural territory that used to represent the hand (FIGURE 4.3) (Ramachandran, Rogers-Ramachandran, & Stewart, 1992). In other words, fol- lowing deafferentation, sensory cortical areas do not go unused; instead, they are taken over by their neighbors (Barinaga, 1992; Flor et al., 1995; Merzenich, 1998). Later in this chapter, we will examine how these changes occur at the level of cells, but here we attend to the bigger picture.
A lthough the cortex that formerly responded to touch on the arm comes to respond to the face, the takeover is not complete—and this means that parts of the brain that it proj- ects to are “expecting” information about the arm. This can lead to perceptual confusion in the form of a phantom sensa- tion. A lmost 150 years ago, S. Weir Mitchell observed Civil War amputees at a hospital in Philadelphia and noted the
curious fact that many of them contended that they could still feel their missing limb (Nathanson, 1988).
What could explain these strange aspects of D.M.’s experi- ence? Her story is not uncommon among patients with amputations—this sort of experience happens to the majority. This is because many square centimeters of the cortex that pre- viously responded to the limb now begin to respond to touch on the face, trunk, or the limb stump (Borsook et al., 1998; Cohen, Bandinelli, Findley, & Hallett, 1991; Merzenich, 1998; Pascual-Leone, Peris, Tormos, Pascual, & Catala, 1996; Yang et al., 1994). As a result, touches along the chin and jawline (represented next to the hand in the homunculus) engender the sensation of the missing hand being touched (FIGURE 4.4).
Over time, the distortion of the body—in which the phantom fingers move closer to the stump—is also a common
04-Eagleman_Chap04.indd 106 02/11/15 3:27 pm
The Brain Dynamically Reorganizes to Match Its Inputs 107
# 158305 Cust: OUP Au: Eagleman Pg. No. 107 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
RESEARCH METHODS: Mapping Out the Brain How did researchers know that the homunculus of the monkeys had changed? They inserted an elec- trode into different parts of the so- matosensory cortex and determined what local neurons responded to. By moving the electrode to different locations and touching the monkeys in different places, they could map out the receptive fields of the neurons— in this case, the area of the body that caused the neurons to fire more (FIGURE 4.5). With current
technology, researchers can use fMRI with monkeys to determine the responses of different brain regions to different areas on the body—but note that this has lower resolution.
For humans, assessing the brain’s map of the body can also be accom- plished with fMRI. But the process is otherwise a little different for the humans. Although it’s possible to use electrodes to measure individual neurons in the motor cortex, it’s rare. But instead there’s a different
opportunity: just ask them where they perceive the touch! In assessing cortical reorganization in the phan- tom limb example above (FIGURE 4.4), researchers asked, “where do you feel this?” while touching a cotton swab to different parts of the partici- pant’s face, neck, and torso. By noting where the man reported the sensation of the touch, they were able to infer how his homunculus had changed.
FIGURE 4.5 To map out the brain’s representation of the body, an electrode is placed in different locations while the animal is touched in different spots. In the case of the monkeys, neurons in the region of the somatosensory cortex that would normally have encoded touch on the arm (a) now responded to touch on the face (b).
Electrode
(a) (b)
2 1
3b
3a
4
2 1
3b
3a
4
report, and the final sensation of the hand is often perceived as being on or even inside the stump (Cronholm, 1951; Henderson & Smyth, 1948; Merzenich et al., 1984; Ramachandran et al., 1992). The same distortion occurs with phantom feet and toes being felt when higher parts of the leg or genitals are touched, and these phantom feet often move toward the stumps of the amputated leg (Ramachandran & Blakeslee, 1998). Quite commonly, enduring pain parallels these changes and is understood as part of the perceptual consequence of cortical reorganization following arm ampu- tation. The magnitude of this phantom limb pain typically correlates with the extent of remodeling recorded in the cortex: the more changes, the more pain (Flor et al., 1995; Karl, Birbaumer, Lutzenberger, Cohen, & Flor, 2001).
The strange phenomenon of phantom limbs illustrates an important point about our perceptions. A lthough, pain, warmth, and touch feel embodied in our limbs, the fact is that they arise in the brain. You can lose a limb without losing sensation in it. (Conversely, if the brain is injured, you can lose sensation in a limb without losing the limb itself). Experiences are built in the circuitry of neurons. Or, more simply, “no brain, no pain.”
Changes to Sensory Input Cortical reorganization does not require an event as drastic as an amputation—it can instead be induced by a temporary
04-Eagleman_Chap04.indd 107 02/11/15 3:27 pm
108 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 108 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The principles of reorganization go beyond the input from the body to the input of the senses more generally. Have you ever heard that blind people have more sensitive touch and hearing? This is because the brain deploys its space and resources according to the signals that come in. Thus, the visual cortex of the congenitally blind becomes tuned to tac- tile and auditory input (FIGURE 4.6) (Elbert & Rockstroh, 2004; Pascual-Leone, Amedi, Fregni, & Merabet, 2005). The perceptual consequence of the cortical takeover is increased sensitivity.
Do we need a lifetime’s experience to reshape our brains so dramatically? The answer seems to be no. W hen people with perfectly functioning visual systems are blindfolded for only two days, their primary visual cortex activates when they perform tasks with their fingers or when they hear tones or words (Pascual-Leone & Hamilton, 2001). Removing the blindfold for just 12 hours reverts the visual cortex so that it responds again only to visual input. The brain’s sudden
change in the sensory input. For example, when a tight pres- sure cuff is fastened to an arm or the nerves from the arm are blocked pharmacologically, within less than an hour, the human brain adjusts to the loss of sensory input by devoting less territory to that part of the body (Weiss, Miltner, Liep- ert, Meissner, & Taub, 2004). In another example, if two fin- gers of the hand of the adult owl monkey are tied together and no longer operate independently of one another, their cortical representation begins to merge into a single area (Clark, A llard, Jenkins, & Merzenich, 1988). In these cases, the brain circuitry adjusts itself to fit the body it is dealing with. The rapidity with which this happens suggests that there does not need to be a large-scale rewiring of these areas, but, instead, that there are connections already in place that are merely unmasked by these changes to sensory input (Clark et al., 1988). This flexible matching to the body plan allows the brain to optimize its allocation of neural resources.
FIGURE 4.6 Cortical reorganization. In this fMRI image, auditory and tactile tasks activate the otherwise unused visual cortex of early blind participants. Brain regions activated more in the blind than in the sighted are shown in the orange–yellow spectrum; areas more active in the sighted than in the blind are shown in blue– green. To see the gyri and sulci (the hills and valleys) of the cortex, the brain has been artificially “inflated” using a computer algorithm. Figure from Renier et al. (2010).
04-Eagleman_Chap04.indd 108 02/11/15 3:27 pm
The Brain Distributes Resources Based on Relevance 109
# 158305 Cust: OUP Au: Eagleman Pg. No. 109 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
trained on two different tasks. The first task required the mon- keys to retrieve small objects via skilled, fine use of the digits. The second was a key-turning task, which required more wrist and forearm use (Nudo, Milliken, Jenkins, & Merzenich, 1996). Then the researchers mapped out how much of the monkeys’ motor cortex was devoted to moving each body part. After training on the first task, the cortical representation for digits progressively usurped more territory whereas the wrist and forearm representation shrank (FIGURE 4.7). In contrast, if the monkeys trained on the key-turning task, the amount of neural territory devoted to the wrist and forearm expanded.
These changes demonstrate that the cortex comes to reflect an animal ’s actions and goals—these become di- rectly reflected in the structure of the brain. We can see these changes in humans as well. For example, learning to play the violin or learning to read Braille both result in in- creased finger representations in cortical maps (Elbert, Pantev, Wienbruch, Rockstroh, & Taub, 1995; Karni et al., 1995). A different study using structural MR I found that when people learned how to juggle, there was a measurable increase in gray matter volume in visual areas of the brain (Dragansk i et al., 2004). Three months after the jugglers quit their new hobby, their gray matter volume shrank back to starting levels. (Note that the volume increases may result from many factors, including increases in the size of cell bodies, the birth of new neurons or glia, or changes in the dendrites, all of which can be expressions of plasticity). By devoting more resources to novel tasks, brains can opti- mize their circuitr y based on the goals in front of them.
Knowing how training can reshape the motor cortex has led to a better approach for patients who are paralyzed after a stroke: get the patient to use the paralyzed limb rather than depend on the good one. In constraint therapy, a function- ing hand or arm is actually bound with a strap device to force the patient to use the weaker one instead. W hy does that help? By trying to use the weaker hand, the patient engages the remaining connections in undamaged areas of the brain. These connections can gradually strengthen until new, func- tional motor control circuits are built. More importantly, restraining the healthy hand prevents its representation from marching in and taking over what few neural resources are left to the disabled hand (Hart, 2005). As in some social circles, the weaker members can survive and even thrive if they are given a little room to succeed.
The Role of Relevance: Gating Plasticity with Neuromodulation From what we’ve covered so far, you might think that prac- ticing an act repeatedly is the key to increasing its cortical representation. A lthough this is part of the story, it is not the whole story: there is a deeper principle at work, and that is whether the actions have relevance to the animal. In this
ability to take over the visual cortex depends on connections from other areas that are present but unused under normal circumstances, an issue to which we will return later.
Beyond increased sensitivity, another consequence of cortical reorganization can be hallucination. As an example, patients whose auditory nerve is severed may sometimes ex- perience tinnitus, a constant ringing in the ears that occurs without any actual auditory input (Lockwood, Salvi, & Burkard, 2002). Similarly, progressive hearing loss can lead to musical and verbal hallucinations (Miller & Crosby, 1979). These symptoms result from cortical reorganization, and they are called auditory phantom sensations (Muhlau et al., 2006).
In short, brains redeploy their territory in the face of changing inputs, a strategy that allows for remarkable adapt- ability. For example, imagine that you were an animal that ended up near waterfalls in the rain forest, where you could no longer hear other environmental sounds. The parts of your brain devoted to hearing would lose territory and other parts of the system would usurp that neural real estate. Like- wise, if you were a fish that evolved to trawl lower depths of the ocean, where shafts of sunlight no longer reached, your visual systems would give up the territory they once com- manded. This plasticity gives brains the capacity to find themselves located in a variety of environments and equipped with a variety of senses—and the neural wiring will flexibly adapt to “wrap itself around” the inputs.
The Brain Distributes Resources Based on Relevance The last section looked at how the brain reorganizes in re- sponse to changes in sensory input. We will now learn how an animal’s behaviors and actions play a critical role in pat- terns of change.
The Role of Behavior Brains appear to employ adaptive coding, which means that they allocate more or less neural activity to any given func- tion depending on the needs of the organism (Schweighofer & Arbib, 1998). In other words, if you decide to make a career change to ornithology, more of your neural resources will become devoted toward learning the subtle differences be- tween birds (wing shape, belly coloration, beak size), whereas previously your neural representation with respect to birds may have been crude, such as, “is that a bird or an airplane?” Your sensory abilities refine themselves as required.
This is true not only of sensory representation, but also of motor representation. In one experiment, monkeys were
04-Eagleman_Chap04.indd 109 02/11/15 3:27 pm
110 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 110 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
NEUROSCIENCE OF EVERYDAY LIFE: Pianists and Violinists have Different Brains Because cortical circuitry can come to reflect the behaviors of the animal, the brains of highly trained musicians become measurably different— in a way that can be di- rectly seen by coarse inspection of the brain’s surface. In 2006, Bangert and Schlaug examined three- dimensional images of many brains, paying careful attention to a region of the motor cortex involved in hand movement (Bangert & Schlaug, 2006). They found that a particular gyrus in this region was shaped dif- ferently for some people than for others, and when the gyrus was puckered up in a particular way, they termed it the Omega Sign (FIGURE 4.8). Independent judges
FIGURE 4.8 Anatomical differences between string players and pianists can be seen with the naked eye.
FIGUE 4.7 Functional mapping of primary motor cortex. When a monkey trains on a task that requires fine-digit manipulation (such as grabbing small objects), the cortical representation of digits expands. Shown here is a functional mapping of the primary motor cortex, demonstrating an expansion of the digit representation (purple) and a shrinkage of the forearm representation (green).
Digit
Wrist/forearm
Digit + wrist/forearm
Proximal
No response
Pre-training Post-training
% t
o ta
l
60 Pre
Post
Pre
Post
Pre Post
80
40
20
0 Digit Wrist/
forearm Digit +
wrist/forearm
04-Eagleman_Chap04.indd 110 02/11/15 3:27 pm
The Brain Distributes Resources Based on Relevance 111
# 158305 Cust: OUP Au: Eagleman Pg. No. 111 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Casamenti, & Pepeu, 1996). These neurons are driven by both rewards and punishments in proportion to their inten- sity (R ichardson & DeLong, 1991). In this sense, they are using both good and bad outcomes as markers that it is time to change the brain.
Electrical stimulation of these cholinergic neurons in- creases plasticity in their target areas, whereas blocking the activity decreases plasticity (Hasselmo, 1995). To illustrate this point, let’s examine how the auditory cortex of adult rats changes after exposure to auditory tones of various frequen- cies. The auditory cortex contains maps like those of the motor cortex, but they are tonotopic rather than somatotopic: they represent adjacent frequencies rather than adjacent body parts. Exposure to a tone does not result in changes to the tonotopic map, but when a particular tone is paired with electrical stimu- lation of cholinergic neurons, the cortical representation for that tone’s frequency massively expands (Kilgard & Merzenich, 1998). In other words, the brain devotes more territory to that tone’s frequency because the presence of the acetylcholine in- dicates that the tone must be important. Parallel results of plasticity-induced effects of acetylcholine have been demon- strated in the visual cortex (Bear & Singer, 1986) and somato- sensory cortex (Sachdev, Lu, Wiley, & Ebner, 1998). Similar changes can be imaged with f MRI in humans in an auditory paradigm: plastic changes are blocked by a pharmacological blocker of acetylcholine (Thiel, Friston, & Dolan, 2002).
section, we will begin to see how the brain can turn plasticity on and off in particular places and at particular times, ac- cording to what is important for the animal’s needs. This ability to allow changes to occur only when something im- portant happens is called gating (think of opening and clos- ing a gate to allow something to pass through). But how does the brain identify that something important has happened and thereby change to encode it?
One way that significance is expressed, biophysically, is through neuromodulatory systems: widely broadcast neural systems that correlate with reward, punishment, and alert- ness. We will learn more about these systems in later chapters. For now, you simply need to know that neuromodulators are diffusely released chemical signals that can gate plasticity such that changes take place only at the appropriate times, instead of each time activity passes through the network. In other words, reorganization of parts of the cortex only occurs when paired with the release of particular neuromodulators (Bakin & Weinberger, 1996).
One particularly important neuromodulator is the neu- rotransmitter acetylcholine (Gu, 2003). Neurons that re- lease acetylcholine are called cholinergic, and these neurons exist mostly in the basal forebrain, a subcortical collection of structures that project to the cortex (FIGURE 4.9). These cholinergic neurons are active when an animal is learning a task, but not once a task is well established (Orsetti,
FIGURE 4.9 Cholinergic pathways in the brain. Of special importance is the nucleus basalis, which transmits acetylcholine broadly throughout the cortex.
Pedunculopontine nucleus and laterodorsal tegmental nucleus
Basal forebrain
Nucleus basalis
Medial septal nucleus and nucleus of diagonal band
compared the brain structures of musicians and nonmusicians and discovered that musicians’ brains reliably showed the Omega Sign, whereas nonmusicians’ brains did
not. Even more strikingly, the type of musician could be distinguished: keyboard players showed a larger Omega Sign in the left hemisphere, whereas string players showed it
more in the right hemisphere. Who knew that what you choose to do with your life can end up reflected in the gross anatomy of your brain?
04-Eagleman_Chap04.indd 111 02/11/15 3:27 pm
112 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 112 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 4.10 Cholinergic transmission disrupted by lesioning the nucleus basalis. After lesioning the nucleus basalis, rats do not have an expansion of the motor map associated with the reaching task and do not improve at the task. Figure adapted from Kilgard (2003).
Reach training
Motor map expands, reaching accuracy improves
Reach training with nucleus basalis lesion
Motor map and reaching accuracy remain unchanged
CASE STUDY: The Government Worker with the Missing Brain A 44-year-old man in France went to his doctor because of mild weak- ness in his left leg. After some basic testing, his doctor sent him for rou- tine neuroimaging to investigate the cause. The brain scan revealed something that no one could have guessed: empty space in much of the volume where his brain should reside (FIGURE 4.11) (Feuillet, Dufour, & Pelletier, 2007). To compound the surprising nature of this finding, the man had lived his entire life with his brain in this compressed state and had shown no obvious problems: he was married with two children and worked a white-collar job as a civil servant. His general IQ of 75 did not prevent performance of normal ev- eryday activities, and his verbal IQ was slightly higher, at 84.
To appreciate how amazing this case is, note that our most bril- liant engineers have no idea how to build machines that are so re- silient to per turbation. Imagine that instead of building a normal digital computer or car, we could
harness the principles of dynamic reorganization to build a computer that could sur vive a truck rolling
over half of it or a car that runs just fine after you tear out half its engine.
FIGURE 4.11 Brain scan of 44-year-old white-collar worker with no obvious behavioral abnormalities. A blockage near the arrow (d) had prevented normal flow of the cerebrospinal fluid since the time he was a child, and his ventricles had filled and expanded as a consequence (known as noncommunicating hydrocephalus). Because of remarkable flexibility in programs of neural development, this did not prevent performance of normal everyday activities. LV = lateral ventricle.
A lthough it is said that practice makes perfect, practicing a task is not sufficient to change the brain in the absence of the plasticity-enhancing powers of the cholinergic neurons. Consider this study an illustration of that principle: two groups of rats were trained in a difficult task of grabbing sugar pellets through a small, high slot. In one group, cholin- ergic neurons in the basal forebrain were pharmacologically destroyed. For the normal rats, two weeks of practice led to a 30% increase in the size of the cortical area devoted to the forepaw movement. In parallel, their motor skills and speed improved. By contrast, for the rats without cholinergic mod- ulation, the same cortical area actually shrank by 22%, and accuracy for reaching the sugar pellet never improved (FIGURE 4.10) (Conner, Culberson, Packowski, Chiba, & Tuszynski, 2003). So we see that the basis of plasticity and behav ioral improvement is not simply the repeated perfor- mance of a task: it also requires neuromodulatory systems to encode the relevance of the task.
04-Eagleman_Chap04.indd 112 02/11/15 3:27 pm
The Brain Uses the Available Tissue 113
# 158305 Cust: OUP Au: Eagleman Pg. No. 113 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The ancient Greeks knew the importance of relevance, observing that the best learning occurs when one is inter- ested—that is, paying attention (Benson, 1990). Experimen- tal work bears out their observations: attending to one stimulus over another can increase plasticity in the brain region of the first and not the second. For example, in one study a monkey was exposed to simultaneous auditory and tactile stimulation. If the demands of the task required him to pay attention to the touch, the somatosensory cortex showed plastic changes whereas his auditory cortex did not. If he in- stead was directed to attend to the auditory stimulus, the op- posite happened (Burton & Sinclair, 2000). This effect of attention is mediated by acetylcholine: when we attend, more acetylcholine is released (Sarter, Bruno, & Turchi, 1999).
As a side note, the targets of cholinergic transmission tend to be broadly distributed, not precise (FIGURE 4.9). So why doesn’t acetylcholine release accidentally cause widespread neural changes? The answer is that cholinergic effects can themselves be modulated by other, inhibitory neurotransmit- ters. In combination, those neurotransmitters can make local areas less plastic (or not plastic at all), so that the changes occur only in the specific areas where they are intended (Sarter et al., 1999). As a result, the release of acetylcholine gates the brain to reconfigure small, specific regions of circuitry.
The Brain Uses the Available Tissue We have so far been examining how the brain allocates its resources when sensory inputs or behavioral outputs change. We now turn to a slightly different issue: what happens when the brain’s available resources change—that is, when dis- ease, surgery, or brain damage lead to less brain tissue?
Maps Adjust Themselves to the Available Brain Tissue There are two possibilities that could explain what happens when the brain’s available resources change: first, the system might leave out the parts of the map corresponding to the missing tissue or, second, the brain might make the same map of the body on a smaller piece of real estate. W hich do you think is the case?
To find out, researchers turned to the frog, in whom nerves from the eye travel directly to the optic tectum (roughly anal- ogous to the visual cortex in mammals) (FIGURE 4.12a). There the nerves plug in retinotopically—that is, nerve fibers from the top of the eye connect to the top of the tectum, the left part of the eye to the left part of the tectum, and so on. Essentially, each fiber coming from the eye appears to have a preassigned address where it plugs into the target. To understand the prin- ciples of plasticity, researchers removed half of the optic tectum during development, before the optic nerves had ar- rived. W hat happened? A full retinotopic map developed on the smaller target area (Udin, 1977). The map was compressed in size, but otherwise arranged normally (FIGURE 4.12b).
In a more dramatic demonstration of the same principle, researchers transplanted a third eye in a tadpole (FIGURE 4.12c). This resulted in an unusual situation in which two sets of optic nerves now had to share the same target area of the tectum. W hat happened? The two eyes shared the territory in alternating stripes, each with its full retinotopic mapping (Constantine-Paton & Law, 1978; Law & Constantine-Paton, 1981). In other words, the retinal fibers once again utilized whatever target area was available. In this case, it was not that one half of the tectum was missing—it was simply being com- peted for by the fibers from the other eye, leaving less total territory available.
FIGURE 4.12 Plasticity in the development of the nervous system. (a) Fibers from the tadpole’s eye map retinotopically onto the tectum. (b) If half the tectum is removed, the complete input fits itself onto the smaller available area. (c) If a third eye is transplanted on one side, the tectum reorganizes to accommodate the additional input. (d) If half the retina is removed, the information from the remaining fibers spreads out to cover the available area of the tectum.
(a) (c)
Radioactive dye
Optic tectum
(b)
Half the tectum removed— retinotopic map squishes
New input added— inputs share the tectum
Half the input removed— input spreads over tectum
(d)
Transplanted eye
04-Eagleman_Chap04.indd 113 02/11/15 3:27 pm
114 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 114 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
The last two experiments demonstrate that maps can compress, and even alternate when necessary, to fit the space available. But what if extra territory is available? Can the maps stretch to fill it? To find out, researchers removed one half of a frog’s retina: now only half of the normal number of optic nerves headed out for the normally sized territory of the optic tectum. W hat happened in this case? The retinotopic map (now coding for only half of the visual space) spread out to utilize the entire tectum (FIGURE 4.12d) ( Attardi & Sperry, 1963).
As we will see later in this chapter, neural maps are determined in part from local competitions at the level of neuronal populations, rather than from a prespecified set of blueprints. As a result of these competitions, whatever cortex is available will get used and filled. From an evolu- tionary point of view, these plastic mechanisms give great flexibility to organisms, allowing a single genetic program to work with innumerable varieties of body types. Remember Matthew S.’s hemispherectomy at the beginning of this chapter? We are starting to understand how his brain could survive losing half its territory: it rewired itself to retain most of the normal function on less than the normal amount of brain territory.
Cortical Reorganization after Brain Damage The properties of cortical wiring are the best hope after brain damage caused by stroke. It is typical for a patient’s symp- toms to look the worst immediately after a stroke because of tissue damage and swelling. W hen tissue damage and swell- ing subside, typically after a few days, there is often a sharp recovery of some function. But that’s when the real work of the brain begins. Over the course of weeks, months, or years, massive cortical reorganization can occur, and functions that were lost can sometimes be regained.
An example of this can often be seen with aphasia, in which language skills are damaged or lost following brain injury. In the majority of people, language is localized to the left hemisphere, and after left hemisphere damage language function is impaired. However, with time, language function will often begin to recover—not because of a healing of the left hemisphere, but in theory because of transfer of language function to the right hemisphere. In one report, two separate patients were found who had left hemisphere strokes fol- lowed by language impairment and eventually (partial) re- covery. But both of these unlucky patients then suffered right-side strokes and showed a worsening of their recovered language, suggesting that the function had transferred to the right hemisphere (Basso, Gardelli, Grassi, & Mariotti, 1989). In recent years, researchers have used f MR I to provide con- firmatory evidence for this theory and map this kind of transfer of functions across hemispheres (Hertz‐Pannier et al., 2002).
But how can one brain region take over the functions of another? Aren’t they in different places, doing different things? How does “visual cortex” stop being visual and assume another role? The answer is that the function of a neuron does not depend on its identity or its location, but instead on its connections. If a neuron is lost, but another neuron manages to get access to the same set of inputs and outputs, it can in time assume the functions of the first. W hen it comes to the functioning of neurons (as sometimes with people), you are who you know.
In the case of the aphasia patients, some of the neurons in the undamaged hemisphere have similar connections to the ones that were lost. With some adjustments over time, they can learn to perform the missing f unction. But if these neurons, too, are lost, it may be more diff icult to f ind such well-connected replacements the second time around. At this point, the aphasia becomes much worse (Basso et al., 1989).
To summarize this section, we have seen that the brain’s maps can change and adjust to meet the available neural ter- ritory. This simple developmental rule allows the brain to build maps that can be stretched and compressed. In a map like this, the hand will be represented near the elbow, which will itself be represented near the shoulder—irrespective of how much or how little territory is available. W hen large parts of the cortex are removed, automatic rewiring sets into place that can often recreate the functions of the original cir- cuitry. These examples demonstrate that the organizational structure of maps in the brain do not require full genetic pre- specification, but instead can unfold naturally with an ani- mal’s development and experience.
A Sensitive Period for Plastic Changes A lthough brains change quite a bit in response to their inter- action with the world, they are not equally plastic at all points in time—instead, they are most plastic during a window of time called the sensitive period. A fter this period has passed, the system becomes more difficult (but not impossible) to change.
A Window of Time to Make Changes An understanding of the sensitive period can direct clinical approaches. For example, if a child is born with strabismus (misaligned eyes, known colloquially as “lazy eyed” or “cross- eyed”), her visual system will not wire up correctly. She will favor one eye and the other eye will develop bad vision, also called amblyopia. In terms of visual function, there is nothing physically wrong with the eye on the impaired side; instead,
04-Eagleman_Chap04.indd 114 02/11/15 3:27 pm
A Sensitive Period for Plastic Changes 115
# 158305 Cust: OUP Au: Eagleman Pg. No. 115 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
CASE STUDY: Danielle, the Feral Child in the Window In July 2005, police in Plant City, Florida, pulled up outside a dilapi- dated house to investigate a claim of child abuse. What they found inside caused them to be physically ill. Danielle Crockett (FIGURE 4.13), a girl of almost 7 years of age, had been locked away in a dark closet for, as far as they could tell, her whole childhood (DeGregory, 2008). She was flecked with fecal matter and cockroaches. Beyond basic suste- nance, it appeared she had never received physical affection or normal conversation and that she had probably never been let out- doors. She was fully incapable of speech. Police, social workers, and psychologists all reported that she appeared to look right through them; she had no glimmer of recog- nition or normal human interaction. She could not chew solid food, did not know how to use a toilet, could not nod yes or no, and by one year later had not mastered use of a sippy cup. After many tests, physicians were able to verify that she had no genetic problems such as cerebral palsy, autism, or Down syndrome. Instead, the normal development of her brain had been derailed by severe social deprivation.
As of 2010, Danielle was show- ing some improvements. She had
learned how to use the toilet, could understand what people said to her, and could make some limited verbal replies. She was attending prekin- dergarten and learning to trace
letters. Despite these promising signs, however, it is unlikely that she will be able to gain much of the ground lost during her tragic first years of life.
FIGURE 4.13 Danielle, a feral child discovered in 2005 in Florida. Although she is a beautiful girl, the expressions and behaviors inherent to normal human interaction did not have a chance to develop properly.
the problem for the child lies in the visual cortex, where the dominant eye outcompetes the misaligned one to take over its territory. If the misalignment is not fi xed at an early age (when the eyes are jockeying for control of visual cortex), the vision in the misaligned eye will not be recoverable (Berman & Murphy, 1981).
Just as with the principles of constraint therapy in the stroke patients we saw earlier, the solution is to cover up the dominant eye to give its weaker partner a chance to acquire
some cortical territory. Once the sensitive period has passed, the patch can be removed. This clinical knowledge stems from studies of animals with normal vision, in which one eye is cov- ered with a patch to assess the effects on vision. If the patching is done while the animal is young, it changes the balance of activity between two eyes and permanently ruins the vision at- tainable by that eye. However, the same eye patch applied later in life has no lasting effect because the sensitive period has passed (Issa, Trachtenberg, Chapman, Zahs, & Stryker, 1999).
04-Eagleman_Chap04.indd 115 02/11/15 3:27 pm
116 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 116 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
L; now the change from one to the other is not able to be de- tected (Kuhl, 2004). The sensitive period for language sounds or phonemes (perceptually distinct units of sound that distinguish one word from another) is earlier than that for vocabulary, which is why we hear accents in those who acquire our language late in life (Kuhl, Conboy, Padden, Nelson, & Pruitt, 2005).
We have seen that the acquisition of language and the de- velopment of vision depend on normal input from the world. More generally, what happens if a child receives none of the proper social input, such as physical touch and social feed- back? W hat impact does severe deprivation have on develop- ment? As seen by the tragic case study of Danielle and other feral children (children raised without normal human inter- action), the ability to learn language, interact socially, walk normally, chew food, and have normal neurodevelopment is limited to the years of young childhood. A fter a certain point, these abilities diminish. The brain needs to experi- ence the proper input to achieve its normal connectivity.
Neuromodulation in Young Brains We have learned that the brain is most spongelike and changeable during the sensitive period. However, this does not mean that brain plasticity is solely the privilege of the young. Neural rewiring is an ongoing process: we can form new ideas, accumulate information, remember people and events, and recover from injury. It simply becomes more dif- ficult to do so. So why do young animals have such plastic brains? W hy does learning become harder with age? W hat is the difference between young and old brains that can ac- count for the drastic change in plasticity?
First, let’s consider the situation from a developmental point of view. Species with more plastic brains require longer periods of helplessness. These longer periods allow greater flexibility. For example, the brains of human babies are enor- mously flexible to learning—but compared to other species, human babies are born with few built-in skills and are unable to survive on their own for a long time (many years!). Human adults, on the other hand, are quite good at specific tasks but less flexible. So we see a trade-off between plasticity and ef- ficiency: as your brain gets good at certain things, it becomes less adaptable to others. W hat is the mechanism behind this?
Recall what we learned earlier about plasticity in the adult brain: attention causes widespread cholinergic release, which allows change in the tissue—and this is counterbalanced lo- cally by inhibition in the areas that should not change. The story appears to be different in the developing brain. Young animals show a generalized plasticity without attentional focus. And this seems to be because young animals have high levels of cholinergic transmitters but not the other inhibitory transmitters, which become available only later (Broide & Leslie, 1999; Gopnik & Schulz, 2004; McKay, Placzek, & Dani, 2007; Torrão & Britto, 2002). As a result, baby brains have a constant flow of cholinergic signaling that enables
The Sensitive Period in Language As an example of a sensitive period, consider the acquisition of a second language. You have probably observed that the age at which a child moves to a new country influences how well she will learn the grammar of the new language ( Johnson & Newport, 1989) (FIGURE 4.14). If the age of arrival is before 7 years old, fluency is as high as that of a native speaker. An immigrant of 8–10 years of age has a more difficult time reaching that level of facility with the language, and if the child is already past 17 when moving to the new country, her proficiency is likely to remain low. This illustrates that a new language is not equally learnable at all time points—instead, the ability to learn a language declines with age.
During the sensitive period, input from the world is vital for proper development. Infants who are born deaf, for ex- ample, will fail to make the proper vocal babbling sounds, even if their vocal cords are perfectly capable of doing so (Oller & Eilers, 1988). Without the auditory input, the rest of the system does not “ bootstrap” itself into language acqui- sition. However, deaf infants with parents who speak sign language will pick up on the ability to express themselves with their hands and will display “manual babbling”—that is, their hands will make resemblances to components of sign language (Petitto & Marentette, 1991).
Note an interesting property of the sensitive period: the opening of some doors often leads to the closing of others. Infants are able to hear all possible sounds of human lan- guages, but with progressive exposure to their mother tongue, they lose the ability to hear foreign sounds. For ex- ample, Japanese infants can easily distinguish R and L sounds: if they hear a steady sound like R R R R LLLLL, they will notice the change in the middle. But as they grow slightly older, the sound structure of their own language causes them to lose the capacity to perceive the difference between R and
FIGURE 4.14 Johnson and Newport’s study demonstrated the relationship between age of arrival in the United States and total score correct on a test of English grammar.
230
240
250
260
270
220
210S co
re o
n t
es t
o f
E n
gl is
h g
ra m
m ar
Age of arrival Native 3–7 8–10 11–15 17–39
04-Eagleman_Chap04.indd 116 02/11/15 3:27 pm
Hardwiring versus World Experience 117
# 158305 Cust: OUP Au: Eagleman Pg. No. 117 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
allowed Sperry to examine a simple question: would the fibers from the eye grow back into their original orientation in the tectum or would they twist around to match the world? He discovered that the fibers reconnected to the tectum in their original pattern, as though each fiber retained its original ad- dress despite the fact that the eye had been turned (FIGURE 4.15c). But because the eye was rotated, the unhappy newt now saw the world upside down and shot its tongue downward when a morsel of food was dangled above it.
These findings led Sperry to conclude that the newt’s optic fibers plug into the tectum in a predefined, addressed manner, the way wires plug into a switchboard with a one-to- one addressing scheme. In his chemoaffinity hypothesis, Sperry imagined that each incoming axon might be matched by a particular molecule expressed by each destination cell in the tectum. Later, on realizing that the genome could not possibly code for that many different address molecules, he proposed that gradients of a smaller number of molecules (some repulsive and some attractive) might do the trick, the idea being that each incoming axon will be tuned for a par- ticular combination of concentrations.
Sperry’s chemoaffinity hypothesis does not require the animal to have any experience with the world: it is experience independent. The axons find their way to their targets based on molecular cues, irrespective of the interactions the animal experiences in the outside world. Sperry’s hypothesis was correct (at least, in many places in the brain), indicating that some general aspects of neural connectivity are prespecified and independent of experience. Chemical cues guide the neurons to the right neighborhood, if not their final targets. So, for example, fibers from your retina will always find their way to the visual thalamus; regardless of your experience with the world, they will not wire to your toes.
Experience Changes the Brain As you’ve no doubt guessed, preprogramming cannot be the whole story because much of this chapter has been about dy- namic rewiring depending on circumstance.
Indeed, even early researchers began to suspect that expe- rience would influence the brain—that is, that brains would not develop exactly the same way every time given the same set of genetic instructions. Two hundred years ago, the physi- ologist J. C. Spurzheim proposed that the brain (as well as the muscles) could be increased in size by exercise, based on his hypothesis that “the blood is carried in greater abundance to the parts which are excited and nutrition is performed by the blood” (Spurzheim, 1815). Almost 60 years later, in 1871, Charles Darwin wondered whether this same basic idea might explain why domestic rabbits had smaller brains than rabbits in the wild: he suggested that the wild rabbits are forced to use their wits and senses more than the domesti- cated ones, and their brain bulk reflects that (Darwin, 1871).
By the 1960s, researchers began to study in earnest whether the brain could change in measurable ways as a direct
generalized plasticity. Their brains are generally flexible, slowly coming into focus all at once like a developing photograph, in- stead of changing a little at a time like a pointillist artist.
To summarize, adult brains employ the cholinergic system to allow plasticity and local inhibition to keep it from happening at unwanted spots. Young brains, on the other hand, are bathed in chemicals that allow for global change, making babies the “research and development” department of the human species (Schulz & Gopnik, 2004). As a result, flexibility and skill trade off within a single human lifetime. We will learn more about consequences of the cholinergic system and synaptic plasticity in Chapter 14, when we learn how drugs of abuse can tap into the neuromodulatory sys- tems and literally rewrite the circuitry.
Hardwiring versus World Experience We’ve shown many examples now in which changes in expe- rience change the brain’s wiring. But is there any influence of genetics at all? To what extent does the brain’s wiring come prespecified by genetics, and to what extent does it only absorb information from the outside world to change its own wiring? In this section, we’ll compare hardwiring to world experience and discover some surprising answers.
Aspects of the Brain Are Preprogrammed Brains do not come into the world as blank slates; instead, they are born pre-equipped with expectations about the world. Consider the birth of a baby wildebeest: moments after drop- ping from the womb, it wobbles to its legs and can clumsily run and dodge. In its environment, it simply doesn’t have time to spend months or years learning how to move around. Even human infants, who develop much more slowly, nonetheless have basic reflexes for grasping and sucking and will mimic an adult sticking out her tongue, a feat that requires the sophisti- cated, preprogrammed translation of vision into motor action (Gopnik & Schulz, 2004). And a good deal of work in lan- guage demonstrates that we come into the world pre-equipped to pick up on it (as we will learn in Chapter 11).
In the 1960s, Roger Sperry set out to determine just how much of an animal’s developmental program is hardwired. He turned to the newt, in which nerve fibers from the eye grow back to the optic tectum and plug in retinotopically, just as we saw with frogs earlier. Sperry wanted to understand why each fiber from the eye appeared to have a preassigned address where it plugs in (FIGURE 4.15a). So he cut the optic nerve of an adult newt and rotated the eyeball upside down (FIGURE 4.15b) (Sperry, 1963). Although a severed optic nerve will not regen- erate in a mammal, it will in an amphibian, and that fact
04-Eagleman_Chap04.indd 117 02/11/15 3:27 pm
118 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 118 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
result of experience. The simplest way to examine the question was to rear rats in different environments—for example, a rich environment packed with toys and running wheels or the de- prived environment of a solitary cage ( Bennett, Diamond, Krech, & Rosenzweig, 1964). The results were clear: the envi- ronment altered the brain structure, and this in turn corre- lated with changes in the animal’s capacity for learning and memory. The rats raised in enriched environments performed better at behavioral tasks and were found at autopsy to have lush, extensively branched dendritic trees (FIGURE 4.16b AND c) (Diamond, 1988). By contrast, the rats raised in the deprived environments were poor learners and had abnormally shrunken neurons (FIGURE 4.16d, e, AND f). This same effect of environment was later demonstrated in birds, monkeys, and other mammals (Rosenzweig & Bennett, 1996), making it clear that experience modulates brain development.
Is this effect of environment measurable in humans? Indeed it is (Diamond, 2001). For example, at a Veterans Administration hospital in California in 1993, the brains of college-educated and high school–educated people were compared at autopsy (B. Jacobs, Schall, & Scheibel, 1993). The examiners found that an area involved in language comprehension (Wernicke’s area, to which we return in Chapter 11) had more elaborated dendrites in the college- educated people.
The same picture holds true at the level of brain struc- tures. For example, one study showed that taxi drivers in
FIGURE 4.16 Neurons in the brain of a rat. (a) A representative neuron in the brain of a rat reared in a normal environment. (b and c) In enriched environments the neurons grow more extensive arborizations. (d, e, and f) In deprived environments the dendrites shrink to the point of total disappearance.
(a)
(b)
(c)
Anterior
Dorsal Dorsal
Retina
Eye rotated
Tectum
Ventral Ventral
Posterior Anterior Posterior
New anterior
New dorsal Optic nerve cut
Dorsal
New ventral Ventral
New posteior
New anterior
New dorsal
New ventral
New posteior
Anterior Posterior
Dorsal
Ventral
Anterior Posterior
Axons regrow and plug into the same tectal destinations as before the rotation
FIGURE 4.15 How the newt’s optic nerve makes its connections. (a) Fibers from the retina maintain their organized layout when they plug into the optic tectum. (b) To determine how the fibers find their destinations, Sperry cut the optic nerve and rotated the eye upside down. When the fibers regrew, they plugged into the tectum in their original pattern, (c) This led Sperry to conclude that the fibers do not find their destinations by visual experience, but instead by preprogrammed signaling.
04-Eagleman_Chap04.indd 118 02/11/15 3:27 pm
Hardwiring versus World Experience 119
# 158305 Cust: OUP Au: Eagleman Pg. No. 119 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
second kitten was bound into a gondola that passively car- ried it so it would have the identical visual stimulation as the first kitten, but would not be controlling its own movements (FIGURE 4.18). The kittens who controlled their own motor
central London, as they gained years of experience on the job, gained volume in their hippocampus, an area involved in spatial mapping (we’ll return to this in more detail in Chapter 9) (Maguire et al., 2000).
Brains Rely on Experience to Unpack Their Programs Correctly The above studies demonstrate that brains reflect the environ- ment to which they are exposed. But we can go one step fur- ther: beyond reflecting the environment, brains require the environment to correctly develop. When the first draft of the Human Genome Project came to completion in 2000, one of the great surprises was that humans have only about 25,000 protein-coding genes (Djebali et al., 2012). How does the mas- sively complicated brain, with its approximately 100 billion neurons, get built from such a small recipe book? Part of the answer turns out to be a clever strategy implemented by the genome: instead of hardwiring everything, a more flexible and efficient strategy is to build a rough draft of the general circuitry required and let world experience refine it. Thus, for humans at birth, the brain is remarkably unfinished, and interaction with the world is necessary to complete it (Ijichi & Ijichi, 2004).
As an example, take the sleep–wake cycle that we will learn about in Chapter 10. This internal clock, known as a circadian rhythm, appears to run on a 24-hour cycle, but if you descend into a cave for several days—where there are no clues to the light and dark cycles of the surface—your circa- dian rhythm drifts, typically between about 21 to 27 hours. This unmasks the fact that the brain’s solution was to build a rather nonexact clock and then pin its period to the solar cycle (Recio, Miguez, Buxton, & Challet, 1997). There is no need to genetically prespecify a perfect clock if the world can help out. So although DNA is often called the secret of life, life’s other secret is that there is no need to encode everything if development can exploit the structure of the outside world.
Armed with this new viewpoint, you should now be able to understand why some of the most common problems of vision—such as the inability to see in depth—develop from imbalances in the pattern of activity delivered to the visual cortex by the two eyes. For example, when kittens are raised with artificial strabismus (the two eyes do not look to the same point in space), the activity from the two eyes is not cor- related, as it would be in a normal kitten. As a result, cells in visual cortex involved in binocular vision do not develop, and the strabismic kittens lack stereo vision (FIGURE 4.17) (Lowel & Singer, 1992). The development of normal visual circuits de- pends on normal visual activity. It is experience dependent.
In a particularly striking experiment in the 1960s, re- searchers raised a group of kittens in total darkness (Held & Hein, 1963). Then, for one hour a day, the kittens were ex- posed to light, and during this time one kitten could move around freely in a cylinder with patterns on the wall. The
FIGURE 4.17 Kittens raised with artificial strabismus. histograms show the number of cells in the kitten’s visual cortex that respond to input from one eye or the other, along an arbitrary scale of 1 (activity is driven by input to the contralateral eye) to 7 (activity is driven by input to the ipsilateral eye). Neurons in the middle of the distribution (around 4) respond to activity in both eyes equally—in other words, they are binocular. In the kitten reared with strabismus, almost none of the neurons develop binocularly.
Neurons are binocular
Almost none of the neurons are binocular
N u
m b
er o
f n
eu ro
n s
1 2 3 4
Contralateral eye
Ipsilateral eye
5 6 7 1 2 3 4
Contralateral eye
Ipsilateral eye
5 6 7
Cross-eyedNormal
N u
m b
er o
f n
eu ro
n s
FIGURE 4.18 Vision only develops correctly when correlated with one’s own movement. In the contraption used by held and hein (1963), one kitten controlled the movement—and thereby learned the relationship between its motor actions and the resulting changes in the visual world (a). The other kitten received the same visual input, but was never the one causing it (b). This second kitten had permanently impaired vision.
(a)
(b)
04-Eagleman_Chap04.indd 119 02/11/15 3:27 pm
120 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 120 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
movements, thus learning the proper relationships between action and the resulting feedback from the world, developed normal vision; the passively stimulated kittens had perma- nently impaired vision. This demonstrates that incoming visual information must be correlated with one’s own body movements to wire up the visual system correctly. W hen the experience is manipulated, the normal, expected experience with the world is undermined.
As a result of findings like these, it has become clear that genetic instructions have general rather than specific roles in the detailed assembly of cortical connections. That is, neuronal networks require interaction with the world for their proper development. The roles of genetic instructions are simply to guide neurons into the right general areas and to provide them with general mechanisms for adjusting their connections with other neurons. The role of genetics is analogous to the hosts of a social event: they invite the guests and ensure they end up in the same place and speak- ing the same language, but they do not specif y who must be- friend whom or what they have to talk about. The result is improvisation rather than scripting and the flexibility to adapt to circumstances.
The Mechanisms of Reorganization Throughout this chapter, we have seen dozens of examples of the brain’s ability to change its own wiring. In Chapter 9 we’ll learn the mechanisms by which plasticity occurs at in- dividual synapses; for now, we’ll examine the bigger picture of changes at the level of populations of neurons.
Neurons Compete for Limited Space Neurons and their processes (axons and dendrites) chroni- cally fight for resources to survive, striving to find useful niches in the circuitry of a brain. If they cannot find a role in the larger society of the nervous system, they retract. De- prived of the growth factors that sustain them, they ulti- mately remove themselves from the conversation altogether, dying out.
One example of this can be seen in the neuromuscular junction, the point where motor neurons in the peripheral nervous system contact muscles (FIGURE 4.19). The neurons release neurotransmitter here to control the contraction of the muscle. In the adult, each muscle fiber is individually controlled by a single motor neuron. However, if you were to look at the situation early in development, you would see that each muscle fiber is innervated by axons from several motor neurons. W hat happens in between to whittle things down to a one-neuron-to-one-muscle-fiber relationship?
FIGURE 4.19 Competition at the neuromuscular junction. Early in development, many axons innervate many muscle fibers. As the muscle matures, competition whittles down the playing field until each muscle fiber is driven by a single axon.
Muscle matures
Immature muscle
Mature muscle
Many axons innervate many muscle fibers.
Each muscle fiber is driven by a single axon.
The answer is competition (Sanes & Lichtman, 1999). Only one neuron can survive on each muscle fiber. They have to find an open niche and chronically defend it.
A lthough the neuromuscular junction is one of the most studied systems in the peripheral ner vous system, let’s turn to an example in the central ner vous system. Ocular dominance columns are alternating stripes in the visual cortex that represent cells responding to signals from either the left or the right eye. During development, axons carr y- ing visual information from the thalamus initially branch widely in the cortex (FIGURE 4.20a) and then segregate into eye-specific patches based on patterns of correlated activity (FIGURE 4.20b). This segregation is activity dependent: if all incoming activity is blocked by an injection of tetrodotoxin in the retinas, the axons in the cortex remain overlapped (FIGURE 4.20c). Under normal circumstances, both eyes carr y the same level of activity. But Hubel and Weisel showed that the territor y controlled by one eye or the other could be dramatically changed by experience: shutting one
04-Eagleman_Chap04.indd 120 02/11/15 3:27 pm
The Mechanisms of Reorganization 121
# 158305 Cust: OUP Au: Eagleman Pg. No. 121 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Competition for Neurotrophins In 1941, a young Italian woman named R ita Levi- Montalcini fled from her native Turin into a small cottage in the coun- try, where she lived in hiding from the Germans and Ital- ians: her life was in danger because she was Jewish. W hile in hiding, she set up a mini laboratory in the cottage and worked to figure out how limbs developed in chick em- bryos (FIGURE 4.21). Her work there led to the discovery of nerve growth factor, and for this work she won the 1968 Nobel Prize ( Levi- Montalcini & A ngeletti, 1968).
W hat she had discovered was the first example of a class of life-preserving chemicals called neurotrophins. These
eye of an animal leads to an expansion of the territor y oc- cupied by fibers from the open eye (FIGURE 4.20d) (Gu, 2003; Hubel & Wiesel, 1962; Wiesel & Hubel, 1963b). Just as at the neuromuscular junction, these characteristics result from competition at the synaptic level. Inputs from the strong eye are retained and strengthened, whereas the inputs from the shut eye are weakened and eventually decay (Wiesel & Hubel, 1963a).
Plasticity from world experience also involves a good deal of pruning (retraction of axonal branches) and cell death. Cells can die in one of two ways: necrosis (in an un- controlled fashion) or apoptosis (in a deliberate, controlled fashion). The controlled process of apoptosis avoids collat- eral damage to neighbors, and it is a common sculpting mechanism in embryonic development. For example, the process of turning a human embryo’s webbed hand into a baby’s clearly defined fingers depends on sculpting away cells, not adding them (Kuida et al., 1998). The same princi- ples may apply to the development of the brain. During de- velopment, 50% more neurons than needed are produced (Low & Cheng, 2006). Massive die-off is standard operating procedure: neurons die because of failure to compete for chemicals provided by targets. Remember the experiment that removed one-half of the frog’s tectum and resulted in a compressed map? Decreasing the available real estate in- creased the die-off: there was simply not enough room to provide for all the neurons, forcing some to go away. A l- though the structure of the map was retained, a smaller number of neurons survived.
FIGURE 4.21 A ganglion of sensory cells from a chick embryo cultured in the (a) absence or (b) presence of nerve growth factor.
(a) (b)
FIGURE 4.20 Ocular dominance columns in primary visual cortex result from competition for space. (a) At 15 days in the cat, the input layer of primary visual cortex has approximately uniform input from the left and right eyes. (b) As the animal matures, the connectivity comes to reflect alternating input from both eyes equally. (c) When retinal activity is blocked, the segregation does not occur. (d) When one of a young animal is patched, the inputs from the weak eye progressively shrink as the strong inputs from the other eye successfully fight for the territory.
(a) 15 day-old cat (b) Normal development (c) Incoming activity blocked (d) One eye covered with a patch
Lateral geniculate nucleus
Axons carrying visual information from the thalamus initially branch widely in the cortex.
Axons segregate into eye- specific patches based on patterns of correlated activity
When activity is blocked at the retina, cortical axons remain overlapped.
Shutting one eye leads to an expansion of the territory occupied by fibers from the open eye.
Right-eye axon
Ocular dominance column
Left-eye axon
Visual cortex
04-Eagleman_Chap04.indd 121 02/11/15 3:27 pm
122 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 122 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
protein factors, secreted by the neuron’s target, allow the survival, development, and function of neurons. Neuro- trophins are the currency over which the neurons and syn- apses compete for real estate; they are what drive neurons to make connections, and the neurotrophins stabilize those connections. Essentially, the rule is that those who are suc- cessful at getting these life-preserving chemicals—which promote growth and survival, guide axons, and stimulate the growth of new synaptic connections—live. Otherwise they try elsewhere, or, if they are unsuccessful any where, they don’t survive.
Neurotrophins work in at least two ways. One is by allow- ing a cell to differentiate into its next stage of development. The other way, at least early in the development of the organ- ism, is by preventing a cell from initiating suicide by apopto- sis. In the intervening years since Levi-Montalcini’s initial discovery, many neurotrophic factors have been discovered (Spedding & Gressens, 2008). Additionally, there are several postulated toxic factors—for example, synaptotoxins are thought to eliminate existing synapses (Zoubine, Ma, Smirnova, Citron, & Festoff, 1996). In one model, axons vie to escape the punishing effects of synaptotoxins by remain- ing active—as soon as they drop below a threshold, the axons are eliminated (Sanes & Lichtman, 1999). Some current re- search in Alzheimer’s disease (Chapter 16) suggests that the amyloid beta molecules that are characteristic of the disease act as synaptotoxins (Klein, 2013). Trophic and toxic mole- cules provide signals that allow neurons to determine whether they should remain at their posts in the brain’s circuitry or remove themselves for the sake of the common good.
We have seen that the constant reorganization of the brain circuitry is underpinned by fierce competition for lim- ited resources in the neural tissue, but how quickly do these changes take place? As we will see, some changes happen rapidly, whereas others happen more slowly.
Inhibitory neuron
Loss of normal input
FIGURE 4.22 Due to disinhibition, the widely spread and previously silent projections from the thalamus begin to play a functional role. As a result, the receptive field of downstream neurons can expand to contain neighboring structures.
Rapid Changes: Unmasking Existing Connections Remember the blindfolded subjects whose visual cortex begins to respond to touch within two days? This rapid time period suggests that the connections were already there (Buonomano & Merzenich, 1998; Pascual-Leone et al., 2005; Pascual-Leone & Hamilton, 2001). A popular model suggests that there are many neural connections that already ex ist, but they are inhibited so that they have no effect, functionally speak ing (Buonomano & Mer- zenich, 1998). As an analog y, imagine a major disruption to your circle of friends. Because of a socially tragic misunder- standing at a party (where ever yone else was acting just as wild as you were), you lose all your closest friends. Sud- denly, your social input is less than what it used to be, and now you begin listening for signals from other friends— those with whom you had a tenuous connection but who never had a chance to command your full attention before. Their voices were squelched by the strong relationships you had with your main friends. Now that these peripheral friends can be heard, you begin to fill out your social life by attending to those weak connections and work ing to strengthen them.
A s you can g uess f rom this ana log y, the mechanism for unmask ing is the release of inhibition that the strong connections had prev iously prov ided. T hat is, when the or ig ina l connections lose their active input—say, be- cause of an anesthetized arm or a blindfolding—fast changes in receptive f ields can resu lt f rom the disinhibi- tion of cover t, ex isting connections f rom other sensor y reg ions of the tha lamus to the cor tex (FIGURE 4.22) (Weiss et a l., 20 0 4).
04-Eagleman_Chap04.indd 122 02/11/15 3:28 pm
Changing the Input Channels 123
# 158305 Cust: OUP Au: Eagleman Pg. No. 123 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 4.23 Growth of new neurites into a region after loss of previous input. (a) Neuron 1 innervates the target; neuron 2 does not. (b) Loss of input to neuron 1 occurs. (c) Neuron 2 projects to target, replacing the input of neuron 1.
(a)
Neuron 1 Target
Neuron 2
Neuron 1
Neuron 1
Target
Neuron 2
Target
Neuron 2
Neuron 1 innervates the target
(c)
Neuron 2 projects into target, replacing the input of Neuron 1
(b)
Input from Neuron 1 goes away
Slow Changes: Growth of New Connections Longer-term changes are thought to involve the growth of axons into new areas and the sprouting of new connections (FIGURE 4.23) (Pascual-Leone et al., 2005). To continue with the social friendship analogy, imagine that you have begun to exchange Facebook messages with those peripheral
friends you never paid much attention to before. With time, given the unexpected room in your social schedule, these friends introduce you to their friends, and you become open to new friendships that you lacked room for before. You seek out and establish brand new connections that stem from more distant social circles. And so it goes with the brain: with enough time, deafferented areas sprout new connec- tions (Darian-Smith & Gilbert, 1994; Florence, Taub, & Kaas, 1998).
Given the t wo time scales of cor tical remodeling , we can now understand a general principle of reorganization: the brain seems to put into place many “silent” connec- tions that are inhibited during ever yday neural conversa- tion—but are there if needed in the f uture. With these, the brain can respond rapidly to changes in input. How- ever, these silent connections are limited in number, and for longer, more w idespread change, a dif ferent approach is used. Essentially, if shor t-term changes (such as un- mask ing) are found to be usef ul to the animal, then long- term changes (such as grow th of new a xons and sprouting of new sy napses) w ill eventually follow (Buonomano & Merzenich, 1998).
Changing the Input Channels The principles of competition and rewiring that we have learned in this chapter have set us up to understand one of the most striking consequences of plasticity: the brain will allow the incorporation of unusual, new inputs (Hawkins & Blakeslee, 2004; Sharma, Angelucci, & Sur, 2000; von Melchner, Pallas, & Sur, 2000), Consider the bionic retinal implant (BR I). The BR I is a gas-permeable patch that is sen- sitive to light and has miniature electrodes that plug into the back of the eye (Dowling, 2008) (FIGURE 4.24). The implant works well in eye diseases in which the photoreceptors at the back of the eye are degenerating, but the retinal ganglion cells (with which the photoreceptors communicate) are per- fectly healthy (more on this in Chapter 5). So the tiny elec- trodes of the BR I replace the normal functions of the photoreceptor sheet and send out their tiny sparks of electri- cal activity. A lthough the signals sent by the implant are not precisely what the rest of the brain is used to, the down- stream processes are able to learn to extract the information they need for vision.
Beyond circumventing a broken peripheral sensor, inputs into the brain can be switched around. In a striking demonstration, scientists at MIT redirected inputs from a ferret’s eye over to its auditory cortex (through the medial geniculate nucleus; FIGURE 4.25). W hat happened to the au- ditory cortex? The visual inputs reorganized it, altering its circuitry to resemble circuitry and connectivity in primary visual cortex (Sharma et al., 2000; von Melchner et al.,
04-Eagleman_Chap04.indd 123 02/11/15 3:28 pm
124 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 124 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
2000). The rewired animals behaviorally interpreted visual input (now going to the auditory cortex) just as they did with normal vision.
One framework for understanding this result is to think of the neocortex as a data-processing engine that accepts
FIGURE 4.24 The bionic retinal implant. A camera mounted in front of the eye sends its video feed to an electrode array at the back of the eye.
Video camera
Antenna
Optic nerve
Retinal implant
Image captured by video camera
1.
Image transmitted to computer (on belt)
2.
Digitized image sent to device behind ear
3.
Device transmits signal to retinal implant
4.
Retinal implant uses optic nerve to transmit image to brain
5.
whatever input is plugged into it and performs the same basic algorithms on all input (Hawkins & Blakeslee, 2004). The inputs will compete for space, and downstream neural popu- lations will learn how to interact with them. Plug a visual data stream into a patch of cortex, and it will become what
FIGURE 4.25 Visual fibers can be rerouted to the auditory cortex, enabling the auditory cortex to “see.”
Inputs from eye are redirected to the auditory cortex.
1.
Visual inputs alter circuitry to resemble the circuitry and connectivity in the primary visual cortex.
2. Rewired animals interpret visual input in the auditory cortex.
3.
Medial geniculate nucleus (MGN)
LGN
Auditory cortex
Visual cortex
04-Eagleman_Chap04.indd 124 02/11/15 3:28 pm
Changing the Input Channels 125
# 158305 Cust: OUP Au: Eagleman Pg. No. 125 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
we think of as a visual cortex; plug an auditory stream into it, and it will become an auditory cortex.
As we have seen so far, the brain has a remarkable capac- ity to reconfigure itself in the face of new inputs, outputs, neural gains, or neural losses. That versatility opens the door to technologies that can deliver information to the brain through unusual sensory channels. For example, what if a blind person used the data stream from a video camera and converted it into sounds in her headphones? Would she eventually be able to see the world by listening? Welcome to the world of sensory substitution experiments, in which a deficient sensory channel is circumvented in favor of other routes to the brain (Lenay, Hanneton, Marque, & Genouel, 2003; Maidenbaum, Abboud, & Amedi, 2013; Poirier, De Volder, & Scheiber, 2007).
W hy does the BrainPort use the tongue, of all places? A lthough we normally think of the tongue as a taste organ, it also has the finest sense of touch on the entire body. It can distinguish stimuli only 1.6 mm apart—roughly twice as fine as a ty pical fingertip (Wilson, Walton, Tyler, & Wil- liams, 2012). This makes the tongue a great site for passing on new k inds of information, even learning to “see.” A grid of electrodes, the size of a postage stamp, zaps the tongue, converting the lattice of video pixels into “pixels” in the mouth (Bach-y-R ita, 2004; Bach-y-R ita, Collins, Saunders,
W hite, & Scadden, 1969). With practice, the tongue learns to interpret the signals that correspond to the visual prop- erties, such as how large an object is, how far away it is, and whether it’s moving in a particular direction. With the BrainPort, blind users can learn to navigate complex ob- stacle courses and throw balls into buckets. For sighted people, the BrainPort can be used to see in the dark . How is any of this possible? Vision isn’t about the eyes. It’s about the brain (K hoo, Seidel, & Zhu, 2012). Other sen- sor y substitution dev ices for the blind convert v ideo streams into patterns of touch on the lower back, sound for the ears, or small electric shocks to the sk in of the fore- head. Similarly, a sensor y substitution dev ice for the deaf uses a vest covered in v ibrator y motors to translate sound into patterns on the sk in (Nov ich & Eagleman, under rev iew). These amazing substitutions are possible only be- cause the brain can dy namically shape itself around what- ever input is presented. It even seems possible that in the near f uture people w ill feed information streams directly into their cortex.
In conclusion, the rerouting of information and the suc- cess of sensory substitution underscores the dynamic plas- ticity of brains. The principles of competition constantly reorganize the circuitry to optimize their representation of the input.
CASE STUDY: The Man Who Climbs with his Tongue Eric Weihenmayer is a mountaineer who has scaled Mount Everest—a feat made even more impressive by the fact that he is blind. As a child, Eric progressively lost his vision to a rare eye disease called retinoschi- sis, and he was rendered entirely blind by the age of 13. But that didn’t slow his ambition to become a climber. Given his condition, it’s cap- tivating to watch Eric scale shear rock faces, holding on to small crev- ices and protrusions. How does he know where to reach next? How does he do it?
Eric climbs with an electrode grid in his mouth called the BrainPort (FIGURE 4.26). The grid delivers little impulses to his tongue that mirror
the visual signals from a camera at- tached to his forehead. Eric reports that he first had to think hard about how the tongue stimulation might translate into edges and shapes. But
he learned, eventually, to recognize the stimulation as direct perception (Levy, 2008). He is now able to use the device for a low-resolution but effec- tive sense of his visual surroundings.
FIGURE 4.26 BrainPort. The BrainPort converts a video feed to corresponding electrical activity on the tongue. With this technology, blind users can come to understand their visual surroundings with high accuracy.
04-Eagleman_Chap04.indd 125 02/11/15 3:28 pm
126 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 126 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
THE BIGGER PICTURE: Adding New Peripherals Note that the plasticity of the brain allows us to think beyond sensory substitution and into the realm of sensory addition. For example, researchers have genetically engineered mice to express color photopigments and thus to have color vision when they normally do not (G. H. Jacobs, Williams, Cahill, & Nathans, 2007). Similarly, a third type of color photorecep- tor has been genetically engi- neered into adult monkeys who normally have only two types— this gave them trichromatic in- stead of dichromatic vision (Mancuso et al., 2009).
Several do-it-yourselfers have begun the movement of sensory addition in humans. One is Neil Harbisson, a man who was born colorblind. In 2004, inspired by the idea of visual-to-auditory transla- tion, Neil attached a device he calls the “eyeborg” to his head (FIGURE 4.27). The device analyzes a video stream and converts the colors to sounds. The sounds are delivered via bone conduction behind the ear.
Now Neil hears colors. He can put his face in front of any colored swatch and tell you what it is; he’s
taken to painting colorful paintings this way (Miah & Rich, 2008). Even better, the eyeborg’s camera detects wavelengths of light beyond the normal spectrum; when translating from colors to sound, he can encode (and come to perceive in the envi- ronment) infrared and ultraviolet, in the way that snakes and bees do.
In the not-so-distant future, we may be able to add new kinds of func- tionality in the brain—for example, by
plugging weather data or stock mar- ket data into the cortex (Hawkins & Blakeslee, 2004). In line with the prin- ciples of plasticity learned in this chapter, the brain should learn how to correlate this with other data and in- corporate it into perception. After all, when new peripherals are plugged into the brain, competition does its job so that useful information wins a voice in the system.
FIGURE 4.27 Sonochromatic scale. (a) Colorblind artist Neil harbisson wears the eyeborg; (b) his “sonochromatic” scale translates colors detected by the camera into output sound frequencies. The inclusion of frequencies for ultraviolet and infrared allows the auditory system to surmount the normal limitations of the visual system.
(a) (b)
Name
Ultraviolet
Violet
Blue
Cyan
Green
Yellow
Orange
Red
Infrared
(invisible)
(invisible)
Over 717.6 Hz
607.5 Hz
573.9 Hz
551.1 Hz
478.4 Hz
462.0 Hz
440.2 Hz
363.8 Hz
Below 363.8 Hz
Color (visual system)
Frequency (auditory system)
Conclusion Returning to Matthew S.’s hemispherectomy at the begin- ning of this chapter, we now have the tools to understand how his brain could sur vive losing half its territor y: it re- wired itself to retain its overall function on less than the normal amount of brain territor y. This was made possible by competition at the level of the synapses and neurons, which allowed both the rapid unmask ing of existing cross- hemispheric connections and, with time, the growth of new
axons and sprouting of new synapses. A ll throughout, his desire to move, to walk and talk, and to be like other chil- dren his age helped to provide the signals of relevance that allowed plasticity in his brain to express itself fully. This process was helped enormously by the fact that Matthew was only 6 years old—still well within the sensitive period and with lots of extra circuitr y to spare. This remarkable property of neural maps to stretch and compress given the available neural real estate expresses a theme that will come up often in this book: the brain is designed quite differently from a conventional digital computer.
04-Eagleman_Chap04.indd 126 02/11/15 3:28 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 127 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Key Terms 127
genome would allow us to understand how an organism un- packs itself. But development requires interaction with the world; the genome is like a highly condensed recipe with the label “ just add world experience to unpack.” However, like all nature-versus-nurture stories, the interplay between experi- ence dependence and experience independence is complex, and in the end what we find is a challenge to the usual belief that flexibility is at variance with genetic control. During de- velopment, a careful balance of hard-coded and plastic mechanisms confers tremendous flexibility, allowing an animal to optimize its coding machinery to its body plan and its salient tasks.
Looking to the future, as we learn more about how brains dynamically reconfigure their own circuitry, this suggests new ways to build computers and, eventually, to design cars, space stations, and even the houses we live in as reconfigu- rable devices. We can build them in such a way that all the details are not prespecified and that they use interaction with the world to complete their own wiring.
Plasticity is found at all levels, from the synapses to whole brain regions. A theme of this chapter is that the constant battle for territory in the brain is a Darwinian competition: each syn- apse, each neuron, each population is fighting for resources such as neurotrophins. As the border wars are fought, the maps are redrawn in such a way that the goals most important to the or- ganism are always reflected in the structure of the brain. If you drop your career to become a violinist, the neural territory de- voted to your left fingers will expand; if you become a microsco- pist, your visual cortex will reorganize itself to give you higher visual resolution for the small details you search. By examining the key principles of plasticity, we have surveyed a wide swath of territory critical for understanding brain rewiring.
One of the overarching mysteries in neuroscience is how brains are built with so few genes. The fact that we have only 25,000 protein-encoding genes leads us to think about the genome like a compressed file, with the cellular machinery with which it interacts as a decompressor (Marcus, 2004). It was originally hoped that understanding the sequence of the
KEY PRINCIPLES
• The very general layout of brain circuitry is pro- grammed through genetic mechanisms. Real- world experiences fine-tune these general circuits into more detailed programs.
• Early in development, neurons and synapses must either compete successfully for growth factors or die.
• Plastic changes use both fast and slow mechanisms.
• The brain wraps itself around useful inputs, open- ing the door to sensory substitution. The plasticity of the nervous system may allow us to develop ad- vanced devices that connect directly to the brain, to expand human abilities.
• The brain dynamically reorganizes its circuitry to adapt to changes in its sensory inputs.
• The brain reorganizes itself to enable changes in its outputs, distributing resources based on relevance as expressed through plasticity-gating neuromodulators.
• Neural pathways adjust themselves to fit into the available brain tissue. A neuron’s function comes from its inputs and outputs, not its identity or loca- tion. Neural circuits with similar patterns of inputs and outputs can perform similar functions.
• There is a sensitive period during which brains are the most plastic: young brains are more flexible and less specialized than older brains.
KEY TERMS
hemispherectomy (p. 104) sensitive period (p. 105)
The Brain Dynamically Reorganizes to Match Its Inputs
homunculus (p. 105) deafferentation (p. 105)
phantom limb (p. 107) tinnitus (p. 109)
The Brain Distributes Resources Based on Relevance
adaptive coding (p. 109) constraint therapy (p. 109)
gating (p. 111) neuromodulators
(p. 111) cholinergic (p. 111) basal forebrain
(p. 111)
04-Eagleman_Chap04.indd 127 02/11/15 3:28 pm
128 PART 1 • ChAPTER 4 Neuroplasticity
# 158305 Cust: OUP Au: Eagleman Pg. No. 128 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
tonotopic (p. 111) somatotopic (p. 111)
The Brain Uses the Available Tissue optic tectum (p. 113) retinotopic (p. 113) aphasia (p. 114)
A Sensitive Period for Plastic Changes
phonemes (p. 116)
Hardwiring versus World Experience
chemoaffinity hypothesis (p. 117) binocular (p. 119)
The Mechanisms of Reorganization ocular dominance columns
(p. 120) pruning (p. 121) necrosis (p. 121) apoptosis (p. 121)
nerve growth factor (p. 121) neurotrophins (p. 121)
Changing the Input Channels bionic retinal implant (p. 123) medial geniculate nucleus
(p. 123) neocortex (p. 124) sensory substitution (p. 125) sensory addition (p. 126)
REVIEW QUESTIONS 1. What mechanisms might have helped Matthew S.
to have normal abilities despite the loss of half of his brain? If you were to lose half your brain, would you be able to adapt in any of the same ways? If so, how? If not, why not?
2. Why does a “phantom limb” sometimes persist after amputation? How is it possible to have sen- sations in a body part that no longer exists? What might be happening in the brain as the phantom limb disappears?
3. How do stroke patients with aphasia regain their use of language? What kinds of changes take place in their brains? If you were their physician, what kinds of treatments would you recommend to help them recover as much function as possible?
4. Why is there a sensitive period for brain func- tions? Do all brain functions have the same sen- sitive period? What might determine the sensitive period for a given brain function?
5. What is different about the adult brain that makes it more difficult to learn? What forms of
plasticity still exist in the adult? If you had to design a treatment to make the adult brain more plastic, what would you try to do?
6. How do genetic factors influence the wiring of the brain? If you had an identical twin, how might your shared genes give rise to similar talents, interests, and personality quirks? How is it that identical twins can also differ in these areas? How could one identical twin develop a brain- wiring disease such as schizophrenia, whereas the other one does not?
7. How do life experiences fine-tune the brain’s circuitry? How much leeway do they have in doing so? Can life experience turn visual cortex into motor cortex? Why or why not?
8. How do neurons compete for survival? Are they like enemy soldiers, battling for victory over one another? Are they like allies fighting for a common cause? Why might a neuron be “will- ing” to die on purpose if it cannot find a useful function?
CRITICAL-THINKING QUESTIONS 1. If you were somehow born with an extra pair of
arms and hands, would your brain be able to make use of them? How might this be possible? How would your brain be different from the brains of other people?
2. How can the brain fit a sensory map into half the number of neurons if necessary? Could it fit the map into only 10% of the original number of neurons? What about 1%? What capabilities might be lost in the process?
04-Eagleman_Chap04.indd 128 02/11/15 3:28 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 129 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical-Thinking Questions 129
4. Imagine inventing a brain-connected device to extend your own abilities or those of someone you know. What kind of device might you build? How would you enable your brain to use it? How long do you believe it will be before we can build such a device in reality?
3. Sea turtle babies can see, walk, swim, and navi- gate immediately after hatching. Why do human babies have so much trouble surviving on their own? What advantages do they gain in compen- sation? Are there advantages for children who learn to walk, talk, or read especially early in life? Could there be disadvantages as well?
04-Eagleman_Chap04.indd 129 02/11/15 3:28 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 130 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
130
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Give at least two examples of sensory
transduction.
• Trace the path of visual information through the anatomy of the visual system.
• Characterize the hierarchical organization of the visual system.
• Explain how stereo vision works.
• Illustrate the results of damage to primary, secondary, and tertiary areas of the visual cortex.
• Contrast the ventral and dorsal visual processing streams.
• Summarize why we do not see what is actually “out there,” but instead the brain’s internal model of the world.
05-Eagleman_Chap05.indd 130 02/11/15 3:32 pm
131
# 158305 Cust: OUP Au: Eagleman Pg. No. 131 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Vision STARTING OUT: Vision Is More Than the Eyes
Visual Perception Anatomy of the Visual System
CASE STUDY: The Bionic Retina
NEUROSCIENCE OF EVERYDAY LIFE: Random-Dot Stereograms
Higher Visual Areas
THE BIGGER PICTURE: Reading the Movies in Our Minds
CASE STUDY: The World in Snapshots
CASE STUDY: The Blind Woman Who Could See, Sort Of
Perception Is Active, Not Passive Vision Relies on Expectations
CHAPTER 5
05-Eagleman_Chap05.indd 131 02/11/15 3:32 pm
132 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 132 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT: Vision Is More Than the Eyes A chemical explosion left Mike May blind at the age of three. The scarring of his corneas prevented light from getting through. Undeterred, Mike grew up to become a family man, a businessman, and a champion blind downhill skier (FIGURE 5.1). Then, when May was in his mid-forties, a new surgical technique emerged that could clear the scarring of his corneas. He signed up, eager to have the possibility of seeing again.
The operation was a success: when the bandages were peeled off, the corneas were clear, allow- ing light to pass through and onto Mike’s retinas. A photographer was there to capture the moment when he gazed into his sons’ faces for the first time—faces that he had only known through touch. It was guar- anteed to be a very special moment.
But it wasn’t. Although light now passed through his eyes, his brain had no ability to interpret the sig- nals racing along the nerves and spattering into his cortex. Mike stared at everything around him in bewilderment. Faces, hallways, doorframes, windows. None of it
made sense to him: he was only experiencing strange visual noise. The pathway from his eyes to his visual cortex was now functioning as it was supposed to, but he couldn’t be said to be seeing in the normal sense (Kurson, 2008).
Mike’s story illustrates that vision doesn’t come for free. Incom- ing electrochemical signals have to be interpreted by the brain, and that happens only with practice. When Mike’s wife drove him home, he
couldn't understand the whizzing cars around them; nor could he tell whether they were going to smash into the looming highway signs. It has required many years of interac- tion with the world to train up his visual system.
Although vision seems effort- less, it isn’t. The electrical storms of neural activity in the pitch- blackness of the skull get turned into direct perception only after practice. In this chapter we’ll learn how.
FIGURE 5.1 Mike May (left). he is guided by his son, Wyndham, while filming The Movement at the Canyons in Utah.
Visual Perception
What Is It Like to See? In the late 19th centur y, the philosopher and physicist Ernst Mach took notice of something that didn’t make sense. If he took several strips of paper—scaled from white to black—and placed them next to each other, an illusion arose (FIGURE 5.2). Each strip, while uniform in color on its
own, when placed next to the others appeared to have uneven shading, look ing slightly lighter on the side adja- cent to the darker color, and slightly darker on the side ad- jacent to the lighter color (Eagleman 2001).
The illusory color changes at the border are called Mach bands; now that you know about these, you’ll notice them elsewhere. For instance, look at a corner where two walls meet: although the paint is a single color, it can look darker on one wall and lighter on the other. A fter we have introduced some of the cell types involved in vision, you’ll know why this illusion occurs. But in the meantime, we want you to get better at observing your own experience of vision. Chances
05-Eagleman_Chap05.indd 132 02/11/15 3:32 pm
Visual Perception 133
# 158305 Cust: OUP Au: Eagleman Pg. No. 133 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
First, it tells us that the visual system is not like a camera lens that simply reads a photographic representation of the world. This is what one might think from watching a cyborg in a Holly wood movie. As the cyborg stomps around some unlucky city, we occasionally see the world from inside his head. There the world is faithfully reproduced as though his perception were a movie camera, with a heads-up computer display superimposed on the scene for him to read.
But who’s doing the reading? W hat inside his head is watching the display?
In reality, perception is nothing like a movie camera or computer display. Instead, perception is an extraordinarily sophisticated construction of the brain. Sensory machinery is confronted at every moment with a barrage of informa- tion, and it is the task of the nervous system to reduce that amount of information to a single, coherent percept, or mental representation of the thing being perceived.
A lthough this chapter is about v ision, keep in mind that the principles we w ill learn here apply to other per- ceptual systems as well (e.g., hearing and touch), as we w ill see in the next chapter. In all cases, the goal of sensor y systems is to attend to information-rich energ y sources to form a usef ul idea about objects and events in the outside world.
Signal Transduction Signals from the outside world—say, a beam of light, a sound, a smell, a touch—are brought into your nervous system by different kinds of sensory receptors. This process of trans- forming an event from the outside world into electrochemi- cal signals inside your nervous system is called sensory transduction. For mammals, this is usually accomplished through pressure-sensitive receptors on the skin, taste buds
are that although Mach bands have always been in front of you, you have missed them until now. But you shouldn’t feel bad. The entire history of painters until the Renaissance failed to notice that mountains look bluer when they’re far- ther away. Once this was pointed out by a careful observer, then it became a standard trick to paint mountains blue. But why had everyone before this missed it entirely, even though the perceptual facts were right in front of them? Do people simply not pay close attention to their own experiences?
Indeed, we are surprisingly poor observers of our own ex- periences: we assume we know what’s out there until it’s proven otherwise. In this chapter we’ll learn to pay close attention to our experiences, similar to Mach’s observation of the illusion with the strips. By doing so, we’ll extract clues about the mas- sive visual machinery that lies behind our experiences.
We’re going to begin with our perceptions— that is, our experience of the sensory world. Perceptions have a long his- tory of interest, perhaps because of their easy access: all we have to do is open our eyes, and the world seems to be there in all its sensory glory. Vision seems so effortless and seam- less that describing it sometimes feels to people as challeng- ing as a fish trying to describe water. Since the fish lives surrounded by water and knows nothing else, it would pre- sumably have a hard time coming to a conceptual under- standing of water, that is, until it experiences air for the first time, say from a bubble rising from a vent on the ocean floor. In the same way, insights in the field of neuroscience are often achieved with the help of visual illusions. Illusions highlight “ bubbles” in our perception, encouraging neuro- scientists to “swim” closer to have a better look. The exam- ples in this chapter will illustrate this point.
For instance, consider the illusion of the rotating snakes (FIGURE 5.3). Nothing on the page is actually moving, yet ob- servers usually report “seeing” the snakes slithering around. W hat does this tell us?
FIGURE 5.2 Mach band illusion. To prove to yourself that each vertical strip in the figure is in fact uniform in brightness, cover up all but one. When that same strip is viewed in context with the others, it appears to be darker on the right side and lighter on the left.
Perceived
Luminance
Physical
FIGURE 5.3 The rotating snakes illusion by Japanese scientist Akiyoshi Kitaoka.
05-Eagleman_Chap05.indd 133 02/11/15 3:32 pm
134 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 134 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Sensory Transduction: The Eye and Its Retina We’ve just talked about light being a wave, which has a wave- length and amplitude. But light is a strange thing. W hen we try to measure it as a wave, it acts like a wave. If we think of it as a particle and try to measure it that way, it acts like a par- ticle and has mass and force. These discrete particles of light are called photons, which enter the eye through a translucent membrane known as the cornea (FIGURE 5.5). It is this cornea that Mike had replaced in the story at the beginning of this chapter.
Light passes through the cornea and is restricted by a ring of colored muscle fibers known as the iris, which controls the amount of light that can enter the eye. The light passes through a hole in the middle of the iris known as the pupil. From there, the light shines through a lens that will focus the image on the retina, which is at the back of the eye.
The retina is a layered structure, composed of five layers of cells. From the side closest to the lens to the side furthest from the lens, those cells are the retinal ganglion cells, which pass information from the eye to the brain; amacrine cells, which allow communication between different parts of the retina; bipolar cells (a type of bipolar neuron), which carry information from the photoreceptors to the retinal ganglion cells; horizontal cells, which also allow communi- cation between adjacent parts of the retina; and, finally, the photoreceptors. Light passes through all of those cells, in that order, when it enters the eye. Photoreceptors capture the photons of light and convert the light into neurochemi- cal activity through a biochemical process known as photo- transduction. In phototransduction, light strikes a pigment molecule, such as rhodopsin, within the photoreceptor, causing it to break into pieces. These pieces act on proteins in the cell to change the resting membrane potential and, thereby, change the neurotransmitter signal the photorecep- tor is releasing. With time, an enzyme puts the pigment mol- ecule back together, and the cell is ready to signal again. The information that results from this process flows out of the
on the tongue, photoreceptors in the eyes, hair cells in the inner ear, stretch receptors in the muscle, and so on. These are our windows to the outside world.
We only possess transducers for limited sorts of informa- tion. An eyeless animal (say, a worm) does not detect what we call visible light, try as it might. Similarly, the fact that we have no sensors for electromagnetic radiation in the kilo- hertz and megahertz range means we do not perceive the radio, television, and cell phone signals that pour through our bodies at all hours. Our windows to the world outside are small (although generally good enough for survival).
One of our most prized sur vival tools is our perception of visible light—as evidenced by the fact that about 30% of the human cortical real estate is devoted to detecting and processing vision, compared with only 8% for touch and 3% for hearing (Grady, 1993). W hat exactly is visible light? Electromagnetic radiation is energ y that travels through space in a wavelike pattern. Various frequency ranges of this energ y are categorized as radio waves, microwaves, in- frared radiation, ultraviolet radiation, X-rays, gamma rays, visible light, and so on (FIGURE 5.4). Our eyes are a sophisti- cated tool that can detect a small portion of this electro- magnetic radiation, wavelengths from about 390 to 750 nm. This frequency range is called the visible spectrum or visible light. It is surprising that we can sur vive with only information from this infinitesimally small range, just one 10-billionth of the entire electromagnetic radiation spec- trum! Other species, such as bees, butterflies, and some birds, see in the ultraviolet frequency range, whereas some snakes pick up information in the infrared range. However, visible light is traditionally categorized as the range humans can detect (FIGURE 5.4).
Anatomy of the Visual System Let’s examine exactly how our visual system senses and pro- cesses information from the outside world.
FIGURE 5.4 The electromagnetic spectrum. The fraction we detect, and thus call “visible light,” is approximately one 10-trillionth of the spectrum.
350 400 500 600 700 750
Amplitude Wavelength
nm
Visible light
Gamma rays X-rays Infrared
Increasing wavelength
Increasing energy
Radio wavesUltra- violet
Radar TV FM AM
0.0001 nm 0.01 nm 10 nm 1000 nm 0.01 cm 1 cm 1 m 100 m
05-Eagleman_Chap05.indd 134 02/11/15 3:33 pm
Anatomy of the Visual System 135
# 158305 Cust: OUP Au: Eagleman Pg. No. 135 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
several layers of curtains to try to see something; the image would be distorted. The second reason has to do with how the photoreceptors are connected to the bipolar and retinal gan- glion cells. Each cone is connected to its own bipolar cell and then to its own retinal ganglion cell. In contrast, many rods connect to a single retinal ganglion cell (Calkins, Schein, Tsukamoto, & Sterling, 1994). For information coming from the cones, this means that when a retinal ganglion cell is acti- vated, the light that caused that activation can come from only one place on the retina. For the information coming from the rods, the stimulus that activates the retinal ganglion cell could come from any one of many (nearby) places. Be- cause of this, we say that the cones have high spatial resolu- tion and that the rods have low spatial resolution.
eye to the brain through those same cells, but in the reverse sequence of the process described just above.
Photoreceptors are of two types: rods and cones, each of which has a distinct job in the retina. Rods are more numer- ous, with 90 million cells, compared with only 4.5 million cone cells in the human eye (Mainster, 2005). Rods are highly sensitive to light and therefore ideal for vision in dim environments. However, they are broadly receptive to a wide range of light frequencies. Because they do not respond se- lectively to a particular frequency of light, they are not more responsive to one color than to another, and they therefore simply detect degrees of light and dark.
Cones, by contrast, are 10 to 100 times less sensitive to light than rods. They are best suited for vision in bright envi- ronments (e.g., during daylight). Cones come in three types, each of which detects a different distribution of light frequen- cies that peaks around red, blue, and green (FIGURE 5.6). Even- tually, the different activity of these three types of cones in response to objects of different colors will be encoded as color by the brain. Cones are more concentrated in your central vision, a region known as the fovea, whereas there are more rods in the periphery. Have you ever noticed on a dark night that it is easier to see a dim star cluster if you don’t look straight at it? By looking off to one side, you are viewing the stars with your rods, which have greater light sensitivity. Cones are also more sensitive to the fine details of the stimu- lus than the rods are. This is for two reasons. First, you may have noticed in FIGURE 5.5 that the fovea looks like a little in- dentation on the surface of the retina. This is because all those overlying layers of cells are pulled aside and the retinal gan- glion cells are literally smaller in this region, to allow light more direct access to the photoreceptors (Hubel, 1995). If this didn’t happen, vision would be like looking through
FIGURE 5.5 The anatomy of the eye and the cellular layers of the retina.
Ganglion cells
Amacrine cell
Bipolar cells
Lens
Iris
Pupil
Conjunctiva
Blind spot
Optic nerve
Retina
Fovea (point of central focus)
Cornea
Sclera
Horizontal cell
Rods and cones
FIGURE 5.6 Spectral sensitivity of photoreceptors. This image shows the sensitivity of rods and cones to the visible wavelengths of light.
500400 600 700450 550 650
Wavelength (nm)
Short (“Blue”)
cone 437 nm
Rod 498 nm
Medium (“Green”)
cone 533 nm
Long (“Red”)
cone 564 nm
R el
at iv
e ab
so rb
an ce
05-Eagleman_Chap05.indd 135 02/11/15 3:33 pm
136 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 136 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 5.7). W hen both the center and the surround are stimulated—as by a large patch of light—the excitation and inhibition cancel out and the neuron responds little. Other retinal ganglion cells do exactly the opposite, responding to light in the surround but not the center (“off-center cell”). For both on- and off-center cells, note that a uniform surface of light does little to activate them; instead, these cells are optimized for detecting differences in light levels from one area to the next—that is, edges.
Because of the center-surround structure, neighboring neurons can achieve contrast enhancement, that is, the am- plification of a difference between the lightness of two sur- faces. And this is the secret behind the first illusion we encountered, Mach bands (FIGURE 5.2). Receptive fields in the uniform regions have a balance between their excitatory cen- ters and inhibitory surrounds. But a receptive field centered on the lighter Mach band gives a stronger response because part of its surround is in the darker area (and so receives less inhibition from the surround). Conversely, the receptive field over the dark band receives more surround inhibition be- cause part of the surround is in the brighter area.
Here’s another example of that concept: five years after Mach had noticed his bands, Ludimar Hermann noticed gray spots between the squares of a grid—and he puzzled over the fact that the spots disappeared when he looked di- rectly at them (Eagleman, 2001) (Hermann grid illusion; FIGURE 5.8). This illusion arises because a retinal ganglion cell receptive field lying at the intersection of the cross has more light falling on its inhibitory surround than a receptive field that lies between two black squares. Consequently, its
The signals from the photoreceptors are relayed through the layers of cells in the retina, described above, and then reach the retinal ganglion cells. The pathway from the retina to the cortex is organized such that each retinal ganglion cell responds to stimulation only in a specific location of the visual scene. The region of visual space in which a stimulus will modulate the activity of a particular neuron is called the receptive field. Retinal ganglion cells have tiny receptive fields that cover visual space, much like you could use many small tiles to completely cover a floor. Retinal ganglion cells have a center-surround structure: a small point of light in the center of the receptive field will maximally activate the cell, whereas a ring of light in the surround (the disk around the center) will inhibit the firing of the cell (“on-center cell,”
FIGURE 5.7 The center-surround receptive field of on-center and off- center retinal ganglion cells.
Light stimulates center region
Light stimulates surround region
No light stimulation
Light stimulates both center and surround regions
On center cell Off center cell
Light on Light on
Light on Light on
Light on Light on
Time
Time
Time
Time
Increased firing rate
Increased firing rate
Decreased firing rate
Decreased firing rate
Baseline firing rateBaseline firing rate
Weak response Weak response
FIGURE 5.8 The Hermann grid illusion arises from center-surround receptive fields in retinal ganglion cells.
+ –
+ –
05-Eagleman_Chap05.indd 136 02/11/15 3:33 pm
Anatomy of the Visual System 137
# 158305 Cust: OUP Au: Eagleman Pg. No. 137 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
CASE STUDY: The Bionic Retina Terry Byland lives near Los Ange- les, California. In 1993 he was diag- nosed with retinitis pigmentosa, a degenerative disorder of his retina. “Aged 37, the last thing you want to hear is that you are going blind— that there’s nothing they can do,” said Byland (Fleming, 2007).
But then he discovered that there was something that could be done, if he was brave enough to try it. In 2004, he became one of the first pa- tients to undergo an experimental procedure: he was implanted with a bionic retinal chip of the type we saw at the end of Chapter 4 (Fleming, 2007). The chip is a tiny device with 16 electrodes that plugs directly into the retina at the back of the eye. A pair of glasses holds a small camera, and the camera wirelessly beams its signals to the chip. The electrodes in the retina give little zaps of electricity to Terry’s surviv- ing retinal ganglion cells, thereby sending signals down the previously silent highways of the optic nerve (FIGURE 5.9). Just as with Michael’s
case, Terry’s nerve functioned just fine: although the photoreceptors had died, Terry’s optic nerve re- mained hungry for signals that it could carry to the brain.
The idea of such a retinal pros- thesis had been considered since at least the 1990s (Dowling, 2009). But no one was certain whether it would work. After all, the language of the retina is extraordinarily complex— and to this day remains largely undeciphered. Would a small elec- tronic chip, speaking the dialect of Silicon Valley instead of Mother Nature, be understood by the rest of the brain—or would its patterns of miniature electrical sparks sound like gibberish?
Under the direction of Mark Humayun at the University of Southern California, the research team implanted the miniature chip in Terry’s eye. With hushed anticipa- tion, the team turned on the elec- trodes individually to test them. Terry reported, “It was amazing to see something. It was like little specks of
light—not even the size of a dime— when they were testing the elec- trodes one by one” (Fleming, 2007).
Terry described his visual expe- riences over the initial days as small constellations of lights. But, as we discovered in Chapter 4, the brain is plastic—and Terry’s visual cortex began to change to extract some- thing out of the signals. After some time, he detected the presence of his 18-year-old son: “I was with my son, walking, the first time—it was the first time I had seen him since he was five years old. I don’t mind saying, there were a few tears wept that day” (Fleming, 2007).
Terry has now had the implant for years, and his brain can make better sense of the signals. Although he cannot recognize the details of indi- vidual faces, he can see them like a “dark shadow” (Fleming, 2007). The resolution of the retinal chip is only 16 pixels; nevertheless, Terry can touch squares presented at random locations on a computer screen, and he is able to cross a city street by
FIGURE 5.9 The retinal implant. In some disorders, photoreceptors of the retina degenerate although the remainder of the visual system remains healthy. In these cases, the diseased photoreceptors can be circumvented by a small chip that speaks directly to the retinal ganglion cells. A head-worn camera captures the scene, and a portable computer worn on the belt converts the input to electrical signals.
Video camera
Video signal
Retinal implant
Retina
Optic nerve
Implant
Video signal
Cable carrying video signal
Ganglion cells Degenerated layer
of photoreceptors
05-Eagleman_Chap05.indd 137 02/11/15 3:33 pm
138 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 138 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
discerning the white lines of the crosswalk (Eagleman, 2016). He proudly reports, “When I’m in my home, or another person’s house, I can go into any room and switch the light on, or see the light coming in through the window. When I am
walking along the street I can avoid low hanging branches—I can see the edges of the branches, so I can avoid them” (Fleming, 2007).
Although incomplete, Terry’s transition from blindness to par- tial sight makes an enormous
difference to his quality of life, and this research has continued with other patients. As Dr. Humayun notes, “It’s amazing, even with 16 pixels, or electrodes, how much our first six subjects have been able to do” (Fildes, 2007).
excitatory center is suppressed, resulting in weaker activity and the perception of less brightness.
Terry’s chip works by sending electrical pulses directly into the retinal ganglion cells. From there, the signals move on toward the brain via the axons of the retinal ganglion cells, which converge to form the optic nerve (FIGURE 5.10). There can be no photoreceptors at the point where the optic nerve leaves the eye, so this is known as the blind spot.
Both of your eyes are facing forward, and both can see an overlapping part of the world. Therefore, we need to describe what you see not as what comes from your left eye or your right eye, but as what is in the left or right visual field, with the dividing line between them directly in front of you. Each of
FIGURE 5.10 The optic nerve and optic chiasm. A xons of the retinal ganglion cells exit the back of the eye and form the optic nerve. At the optic chiasm, the output from the nasal hemiretina crosses over to the opposite side of the brain, while the information from the temporal hemiretina remains uncrossed. After the chiasm, the nerve bundles carry information about the right or left visual hemifield rather than the right or left eye.
Right visual fieldLeft vi sual field
Layers 1 and 2: the magnocellular pathway.
Information from eyes
Layers 3 through 6: the parvocellular pathway.
Nasal hemi- retina
Nasal hemi- retina
Temporal hemiretina
LGN layers
Temporal hemiretina
Optic chiasm
Optic nerve
Optic radiation
Primary visual cortex
Lateral geniculate nucleus
1
2 3
4 5
6
your retinas is divided into two halves by an imaginary line running vertically through the eye and located at the bound- ary between the right and left visual fields. The half closest to your nose is called the nasal hemiretina, and the half furthest from your nose is called the temporal hemiretina.
The optic nerve conducts all of the information from your retina but keeps track of where in the retina the information originated. As you can see in the figure, the optic nerves from the left and the right eye come together at the optic chiasm, where half the fibers from the right eye and half the fibers from the left eye cross over (chiasm means “crossing”). Spe- cifically, those signals from the right eye that carry informa- tion from the right visual field (the right nasal hemiretina)
05-Eagleman_Chap05.indd 138 02/11/15 3:33 pm
Anatomy of the Visual System 139
# 158305 Cust: OUP Au: Eagleman Pg. No. 139 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Throughout the central nervous system, the brain needs a way to organize the information it receives from the outside world. In many parts of the brain, this is done by topographic organization, which means that there is some orderly map- ping from the external world to the internal representation in the brain. We saw the sensory and motor homunculi in Chapter 4 and will see this principle at work again in Chapter 6. Within V1, this topographic organization takes the form of retinotopic organization, meaning that each neuron re- sponds to a particular part of the visual field, and neighbor- ing neurons respond to neighboring parts of the visual field.
The response characteristics of V1 neurons were discov- ered by accident when David Hubel and Torsten Weisel were recording from a V1 neuron in a cat (Hubel, 1995). They were using a 35-mm projector to show various visual stimuli to a cat, but nothing was making the V1 neuron respond. Then, as they removed one of the slides from the projector, the neuron sud- denly responded! As they put the slide back into place, the neuron responded again. After some investigation, they real- ized that this neuron was responding to the edge of the slide. By moving their electrode forward slightly, they were able to record from a different neuron, which responded to edges at a slightly different angle. They realized that different neurons in V1 responded to edges at different degrees of tilt: some vertical, some at 10 degrees, and so on. These neurons in V1 are orienta- tion tuned, meaning that different orientations of an edge (or simply a line) will maximally activate different neurons. The tuning of a cell refers to the stimuli that activate it (FIGURE 5.12).
In addition to discovering what maximally activates V1 cells, Hubel and Wiesel discovered that the cells were com- posed of two different types: simple and complex cells (Hubel & Wiesel, 1968). Simple cells respond to a line at a preferred orientation and particular location in the receptive
cross over (where they will be processed in the left hemi- sphere), whereas those fibers from the left eye that carry in- formation about the left visual field (the left nasal hemiretina) cross over (where they will be processed in the right hemi- sphere). Information from the temporal hemiretinas remains uncrossed and projects to the ipsilateral side of the brain be- cause that information originated in the opposite (contralat- eral) visual field. The consequence is that the entire right visual field is handled by the left hemisphere (irrespective of whether picked up by your right or left eye), whereas the left visual field is administered by the right hemisphere. A fter the chiasm, once the fibers have been sorted into the two visual fields, the nerve bundles are now called the optic tracts.
Path to the Visual Cortex: The Lateral Geniculate Nucleus From the retinal ganglion cell axons, electrical signals carry- ing information about the visual scene are next processed by a portion of the thalamus known as the lateral geniculate nucleus (LGN). As you recall, the retinal ganglion cells as- sociated with the fovea receive their input from the cones and are smaller than the retinal ganglion cells in other parts of the retina. Based on their size, they are referred to as the parvocellular retinal ganglion cells, whereas the others, which are bigger and receive their input from the rods, are known as magnocellular retinal ganglion cells. These two different pathways project to different parts of the LGN, which is composed of six layers in human brains: two receive input from the magnocellular retinal ganglion cells and four receive input from the parvocellular retinal ganglion cells. The magnocellular layers process information from the rods, which carry information about depth, brightness, and move- ment. The parvocellular layers process information from the cones about fine details of the visual scene including form and color. Neurons of the LGN have receptive fields similar to those of the retinal ganglion cells: they are slightly larger, but still with a center-surround organization.
From the LGN, visual information travels via axons known as the optic radiations to the primary visual cortex.
The Visual Cortex A xons carrying information from the LGN connect to the primary visual cortex (V1), (FIGURE 5.11). Because the axons are myelinated and most axons within the cortex are not, the massive bundle of inputs is visibly different from the rest of the cortex, making a white stripe visible to the naked eye in layer 4 of cortex. (Remember from Chapter 2 that the cortex is divided into six layers. Layer 4 is the layer that generally receives sensory inputs from the thalamus.) For this reason, you will often hear the primary visual cortex referred to as striate cortex (Latin stria = stripe).
FIGURE 5.11 Pathway to the primary visual cortex, also known as the striate cortex.
Optic nerve
LGN
Optic radiation
Primary visual cortex
05-Eagleman_Chap05.indd 139 02/11/15 3:33 pm
140 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 140 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
processing at the lower levels. The brain also makes use of parallel processing, analyzing different kinds of information simultaneously in different locations. We’ll return to these topics in more detail later in the chapter.
Following a series of experiments, researchers realized that there was a two-dimensional grid of neurons within V1 (Smith, Singh, & Greenlee, 2000). Like the rest of the cortex, V1 is organized into columns, with all of the cells within a given column performing similar functions. A long one axis of this grid, the cells differ in their orientation sensitivity. A long the other axis, they differ in the source of their input: one set of columns gets its input from the left eye, and the next set of columns gets its input from the right eye. These columns are known as ocular dominance columns. Within this grid of ocular dominance columns and orientation- tuned columns are clusters of cells known as blobs. (Yes, that really is their official name.) These blobs are important for processing the sensory input relating to color. Together, the ocular dominance columns from the left eye and the right eye, the orientation-tuned columns representing a full
field, whereas complex cells respond to a line of the preferred orientation at any location in the receptive field (FIGURE 5.13).
There are two conceptual points to pay attention to here. First, V1 neurons process more diverse information than LGN cells: they respond to edges of specific orientations rather than simply light and dark spots. This gives us our first glimpse of the hierarchy: as we move farther into the visual system, neurons respond to more abstract stimulus charac- teristics. Second, successive stages of the hierarchy are built from parts of earlier stages: LGN cells gather information from retinal ganglion cells, and simple cell responses in V1 are constructed from the input they receive from several LGN cells, and the response of complex cells is built by com- bining the input they receive from several simple cells (FIGURE 5.14). This concept of a hierarchy is fundamental to the processing of sensory information, and in the next chap- ter we will see the same principle at work in the auditory and somatosensory cortex. A lthough there is a hierarchy, the connections within this hierarchy are reciprocal, meaning that higher levels project back down to lower levels, influencing
FIGURE 5.12 Orientation tuning in V1 neurons. (a) Particular regions in a neuron’s receptive field respond with excitation or inhibition to stimulation. (b) As a result, each neuron can be maximally activated by a particular visual orientation. (c) The response of a neuron to different orientations of a stimulus can be measured—the resultant graph is known as the neuron’s tuning curve. (a) (b) (c)
– –
– –
–
– –
–
–
– –
–
–
–
–
–
–
+ + + + + + +
Stimulus
Response
Time
10
0
20
30 (1)
(2)
(3)
Im p
u ls
es /s
ec
OffOn Inhibitory region
Excitatory region
Orientation (degrees)
0 30˚ 90˚30˚90˚
FIGURE 5.13 Response of simple cells and complex cells to stimulation. (a) The dark shaded area represents a neuron’s receptive field, and the yellow line shows where a bar of light is presented. (b) Simple cells respond maximally to a bar at a preferred orientation in a specific location, while complex cells (c) are orientation-selective but location-insensitive. (a) (b) Response of simple cells (c) Response of complex cells
Receptive field
Stimulation
Bar On
Bar Off
Bar On
Bar Off
Time Time
05-Eagleman_Chap05.indd 140 02/11/15 3:33 pm
Anatomy of the Visual System 141
# 158305 Cust: OUP Au: Eagleman Pg. No. 141 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
rotation of orientations, and the blobs are called a hypercol- umn. This hypercolumn contains all of the information that relates to a single location in the visual field.
Two Eyes Are Better Than One: Stereo Vision Primary visual cortex is the first stage in the system where information from both eyes comes together for the first time. As we saw in the previous sections, signals from the two eyes remain separated in the LGN and all the way to V1.
Once in V1, the visual system can gain something clever by combining the information from the two eyes. But what does it gain, exactly?
To answer this, take a close look at a pen in your hand. When you close one eye, you see the pen lined up against cer- tain things in the background, and when you close the other eye, you see the pen lined up against a slightly different back- ground. This seems like it should cause problems when both eyes are open. However, it turns out that the visual system constructs a three-dimensional view of the world by taking ad- vantage of this disparity—the difference between the visual images that each eye perceives because of the slightly different angles from which each eye views the world. Because your eyes do not occupy the same position in space, but instead are offset by about 6 centimeters, the visual system can gather useful in- formation from two different points of view. The visual system has evolved sophisticated methods to compute the best three- dimensional answer from two flat images on your retinas.
Consider FIGURE 5.15, which shows two side-by-side pho- tographs taken from a few inches apart. If you cross your eyes, the two photos will drift and eventually lock together to produce a single image. There’s something special about this new image: it appears genuinely three-dimensional. This simple demonstration of stereoscopy unmasks some- thing profound: you only need to feed the system the correct inputs, and it constructs a convincing reality for you.
FIGURE 5.14 Building successively richer layers of processing from simple parts. (a) When several LGN neurons converge on a V1 simple cell, the new receptive field can be tuned to more than spots—in this case, it becomes tuned to oriented lines. (b) When several simple cells converge onto a complex cell, that neuron can respond to the preferred orientation in many locations.
(a)
(b)
LGN neurons
V1 simple cell
Complex cell
Simple cells
–
–
– –
– –+
+ +
+
+ +
+ +
– –
– –
+ +
+ +
– –
– –
+ +
+ +
– –
– –
FIGURE 5.15 Stereo vision. The two eyes receive slightly different images. Because light from the two subjects falls on different parts of the two retinas, (a) the left eye sees one image of the world and (b) the right sees another. This slight informational difference from the two eyes does not confuse the visual system. Instead, it uses the differences to extract depth information.(a) (b)
05-Eagleman_Chap05.indd 141 02/11/15 3:33 pm
142 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 142 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
NEUROSCIENCE OF EVERYDAY LIFE: Random-Dot Stereograms The amazing capacity of the visual system to take two slightly differ- ent images and match them up led some researchers to ask a simple question: does stereo vision rely on object recognition? In other words, does your visual cortex have to recognize the images above as street scenes before it puts the right- and left-eye images to- gether? The answer came as a surprise to many: recognition is not necessar y for stereo matching. To elegantly prove this point, psy- chologist Bela Julesz created random-dot stereograms to dem- onstrate that the percept of depth can be generated from disparity alone. In his stimuli (FIGURE 5.16), there is nothing recognizable to the naked eye—and yet, when you fuse the images together, a shape in depth is clearly seen (Julesz,
1986). These simple images dem- onstrate that depth computation from disparity is low level. One only needs to trick the system with
appropriate input, and it automati- cally computes its answer—and we experience depth on the com- pletely flat page.
FIGURE 5.16 Depth information from disparity. These images can be viewed by crossing your eyes until the images overlap. If you find that difficult, try holding a long envelope down the middle of the visual field (right between your eyes), such that each eye sees only one side of the image. These random-dot images demonstrate that object recognition is not necessary for perceiving depth from differences between the eyes.
Higher Visual Areas We’ve traveled from the retina to the primary visual cortex and are now ready to begin the real story of vision.
Secondary and Tertiary Visual Cortex: Processing Becomes More Complex V1 feeds its information into a neighboring area known as secondary visual cortex (V2) (FIGURE 5.17). V2 receives direct connections from V1, and (like V1) it is laid out in a map of the visual world: it is retinotopic. Neurons in V2 have slightly larger receptive fields than in V1 (Gattass, Gross, & Sandell, 1981). Given the lessons we learned about hierarchy, it should come as no surprise that V2 receptive fields are built from re- ceptive field information coming from V1. Most neurons here are still tuned to relatively simple stimuli (for example, orien- tation and color), but we can also detect a slight increase in
perceptual complexity as we continue up the hierarchy: for example, neurons in V1 do not respond to the orientation of illusory lines, but neurons in V2 do (Qiu & von der Heydt, 2005). Consider a Kanizsa figure (FIGURE 5.18): the square in the middle does not actually exist, yet observers usually report “seeing” the square there, as well as differences in brightness outside and inside its illusory edges (Sterman, 1994). Many neurons in V2 show an active response if their receptive fields are positioned where the line would be—in other words, they respond as though a real line were actually there. Only half as many cells in V1 do this, and their response is much weaker (Lee & Nguyen, 2001; Zeki, 1996). The response of V2 neu- rons, therefore, is slightly closer to our final perception of the visual scene, rather than simply a reflection of the world.
V2 projects to several other areas, which are collectively called tertiary visual cortex and often summarized as con- taining areas V3, V4, and V5. Scientists disagree as to what should be included in tertiary cortex, but that’s immaterial to the main point: as we move past secondary areas, we are now climbing to higher levels of the hierarchy. Cells begin to respond to more and more abstract stimuli in their receptive
05-Eagleman_Chap05.indd 142 02/11/15 3:33 pm
Higher Visual Areas 143
# 158305 Cust: OUP Au: Eagleman Pg. No. 143 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 5.17 Lateral and medial view of the brain, highlighting higher visual areas, including V2, V3, V4, and V5/MT.
Medial temporal cortex (MT/V5)
Ventral stream
Temporal lobe
Fusiform gyrus
Temporal lobe
Medial superior temporal cortex (MST)
V1
V4
V2
V3
Parietal lobe
Dorsal stream
Parietal lobe
V1
V4
V2
V3
fields—such as houses, faces, and movement, as we will see in a moment.
The plot thickens here. Remember how the information from the cones and rods stayed segregated in the LGN in the parvocellular and magnocellular pathways? That separation of function holds true as one moves up the hierarchy. Even as visual information becomes more abstracted, the informa- tion moves in two distinct processing “streams” (Mishkin & Ungerleider, 1982). One, the ventral stream, deciphers what objects are—in other words, how to identify and categorize them. The other, the dorsal stream, is focused on where ob- jects are and how to interact with them. We meet these two streams in more detail now.
Ventral Stream: What an Object Is From the parvocellular cells in LGN through V1, visual in- formation flows through V2 and V4 into the inferior temporal
FIGURE 5.18 Kanizsa figure. Although the four pie shapes are separate objects, the visual system fills in illusory lines to form a square.
lobe (known as inferotemporal cortex). This is the ventral stream of information processing, known sometimes as the “what pathway” (Rauschecker & Tian, 2000).
As information progresses from posterior regions (near the back of the head) toward the anterior tip of the temporal lobe, neurons go from encoding features (such as lines, curves, and angles) (Hubel & Wiesel, 1968) to specific ob- jects (faces, cars, logos, and so on) (Martin, Wiggs, Unger- leider, & Haxby, 1996; Martin, Wiggs, & Weisberg, 1997). Specifically, as we saw above, V1 neurons have small recep- tive fields and encode simple features of visual stimuli, such as oriented lines. In V2, the size of receptive fields increases by about threefold. Neurons in V2 respond to more complex features of visual stimuli, such as curves, angles, and the bor- ders between textures (Foster, Gaska, Nagler, & Pollen, 1985; Martin et al., 1996, 1997). In V4, receptive fields are even larger, and the stimuli they respond to are more com- plex, such as gratings and line crossings (Gattass, Sousa, & Gross, 1988). Each of these areas (V1, V2, V4) is laid out in a retinotopic map, representing the outside world in the brain (Dougherty et al., 2003).
The complexity of the shapes encoded increases as we pass from V4 into inferior temporal (IT) regions, and it con- tinues to increase as we move from posterior IT cells toward anterior IT cells (FIGURE 5.19) (Kobatake & Tanaka, 1994). By the top of the hierarchy, we find IT neurons selective for complex shapes such as tools and animals (Hung, Kreiman, Poggio, & DiCarlo, 2005; Kanwisher, McDermott, & Chun, 1997). One part of the IT lobe, the fusiform gyrus, seems to be selective for recognizing faces (Kanwisher et al., 1997).
The receptive fields of IT neurons are much larger than those found in early visual cortex, which means they can re- spond to relevant stimuli almost any where in the visual field and gives them the property that they are not focused on where exactly an object is—instead they are focused on what it is. In other words, changes to the object itself (such as a change in shape) will change the firing rate of IT neurons,
05-Eagleman_Chap05.indd 143 02/11/15 3:33 pm
144 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 144 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
objects, either Barack Obama’s face or a Tesla sports car. In sparse coding, a small number of neurons would become active in response to a specific visual stimulus—say, a small cluster of cells for Obama and a completely different cluster for the Tesla. In population coding, most neurons in ventral visual cortex would provide some response when shown either stimulus, but they would fire to different degrees (FIGURE 5.21).
As it turns out, different parts of visual cortex use differ- ent coding schemes. Face recognition, for example, seems to be highly specific to a small number of neurons, at least for famous faces and for people you know well. This is tested by showing a great number of faces to a monkey while record- ing from a single, face-sensitive neuron—and the result is often that the neuron will fire to one face and one face only, not to any of the rest. The same sparse coding has been found with hands, bodies, and letter strings as well.
However, for many other stimulus types—say, houses, cityscapes, or the general shape of an object, population coding seems to be in effect: many neurons are involved at varying levels of response rather than a few at binary (all-or- none) responses (Pasupathy & Connor, 2002).
W hy do some stimuli seem to be sparsely encoded whereas others are population coded? The answer seems to pivot on familiarity with the stimuli being represented: the more familiar the stimuli, the sharper the representation of individual neurons, the sparser the encoding, and the more clustered the neurons become that represent the object. Thus, expertise in a category begets sparse, clustered popula- tions devoted to that kind of object (Op de Beeck, Baker, R indler, & Kanwisher, 2005). Presumably, we all possess small neural populations in our ventral visual stream that are specialized to, say, cell phones—something our ancestors never had in their temporal lobes.
To summarize our tour of the ventral pathway, neurons in early visual areas faithfully encode simple properties of objects. As processing moves in a hierarchy toward the ante- rior part of the inferotemporal cortex, neurons have larger receptive fields and encode an increasingly abstract form of the stimulus—that is, local characteristics (including posi- tion and size) become less relevant. For example, when a neuron in the fusiform face area (an area of inferotemporal cortex specialized for faces) is presented with a face built of
FIGURE 5.20 Position and size invariance. With large receptive fields, neurons in IT are not focused so much upon where an object is nor on the size of the object. Instead, the response is tied to the object’s identity. Yellow bars represent when the stimulus is present.
Fi ri
n g
ra te
FIGURE 5.19 The ventral stream in inferior temporal cortex. The inferior temporal cortex is subdivided into several areas, and the complexity of information processed increases from the posterior inferior temporal cortex (PIT) to the anterior inferior temporal cortex (AIT). Each of these areas has a full representation, or map, of the visual world. In the progression from posterior to anterior, the response of neurons evolves from specific visual features to generalized understanding of objects.
Anterior inferior temporal cortex (AIT)
Posterior inferior temporal cortex (PIT)
Early visual cortex
Feature recognition
Object recognition
Feature conjunctions
“Wh at”
but a change in the position or size of the object will not (FIGURE 5.20) (Sary, Vogels, & Orban, 1993). A response that remains the same irrespective of the position or size (this is called position invariance or size invariance) is critical to recognizing an object in different contexts: you want to rec- ognize your mother’s face no matter where it is or how close or far she’s standing from you.
Given that neurons in IT show size- and position- invariant responses, it is not surprising that these neurons are no longer organized in retinotopic maps (as they were in earlier stages of visual cortex) (Desimone, A lbright, Gross, & Bruce, 1984). So how is the information encoded at these later stages? Generally, there are two strategies for how the brain can encode information in visual cortex: sparse coding and population coding (review by Reddy & Kan- wisher, 2006). Say you were displaying one of two visual
05-Eagleman_Chap05.indd 144 02/11/15 3:33 pm
Higher Visual Areas 145
# 158305 Cust: OUP Au: Eagleman Pg. No. 145 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 5.21 An illustration of sparse versus population coding schemes. (a) Where sparse coding is found in the brain, small numbers of responsive neurons usually lie in close proximity to one another. (b) In population coding, responsive neurons tend to be more spread out. The brain here is viewed from underneath to see the inferior temporal lobes. Figure from Reddy and Kanwisher (2006).
Response to object A
Response to object B
(a) Clustered (b) Distributed
Fi ri
n g
re sp
o n
se
Number of neurons
Sparse Population
1 2 3 4 … N
Fi ri
n g
re sp
o n
se
Number of neurons
1 2 3 4 … N
THE BIGGER PICTURE: Reading the Movies in Our Minds When you watch a changing visual scene, the patterns of activity are constantly shifting. So could you record the changing pattern across the cortex and predict what a person is seeing? It sounds like science fic- tion, but in September 2011 it became reality.
Researchers at Berkeley recorded from their own brains while watching hours and hours of Hollywood movie trailers on YouTube (Nishimoto et al., 2011). They then used a computer to find correla tions between the prop- erties of each video and the patterns of fMRI activity induced in the visual
cortex. From this, the computer “learned” to recognize what people were seeing, based on the patterns of activity in their brains. Next, new subjects watched a new set of videos. Without knowing what the subjects had watched, the researchers tried to reconstruct what those videos looked like, based only on the measured neural activity. After measuring the activity, their model chose the top 100 videos that best matched the activity patterns, and then it averaged those together for its best guess of what the new video must have looked like. The match was stunningly good.
These results suppor t the stor y that we have been building so far: the visual stimulus is broken down and processed by the brain, re- sulting in cer tain types of brain activity. Working in reverse, re- searchers can look at the activity to make predictions about what the initial stimulus must have been. This technology for decoding and reconstructing people’s dynamic visual experiences may be used, someday, to read out one’s dreams or even the visual imaginings of a coma patient.
05-Eagleman_Chap05.indd 145 02/11/15 3:33 pm
146 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 146 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
spatial relationships between objects as well as the coordina- tion of one’s own body in space (Goodale & Milner, 1992).
Because such a major part of the dorsal stream involves the detection of motion, we will begin there, then turn to issues of attention, and finally investigate the effects of damage to the dorsal stream.
One of the oldest visual illusions in neuroscience dates back to Aristotle, who noticed something after intently watch- ing a horse stuck in a flowing river (Mather, Verstraten, & Anstis, 1998). W hen Aristotle looked at the riverbank, it ap- peared that everything was drifting the opposite direction from the river’s flow. W hat Aristotle experienced is now called the motion aftereffect: following exposure to a visual field moving in one direction, other things appear to drift the other way (Tootell et al., 1995). This striking perception is also known colloquially as the waterfall illusion, so named be- cause staring at a waterfall is an efficient way to induce the af- tereffect: when you look away from the waterfall, the rocks and dirt and trees appear to be levitating upward. Perhaps the strangest part of the illusion is that things appear to move, but with no change in their position—a physical impossibility. The rocks don’t end up any higher than where they started.
How can it seem as though motion occurs even though nothing in the outside world changes position? The answer is that the inside of the brain is not equivalent to a television screen (Dennett, 1992), and so it can easily handle motion with no change in position. In fact, there are many illusions of motion with no change in position. Recall the rotating snakes illusion from FIGURE 5.3 at the beginning of this chap- ter. Similarly, FIGURE 5.24 demonstrates that static images can
sketched lines, paint strokes, or pieces of fruit, it responds to the faceness of the object (FIGURE 5.22).
Dorsal Stream: how to Interact with the World Let’s now return to V1 and V2 to examine the other process- ing pathway, the dorsal stream. This path relates less to what an object is and instead processes where it is in space. It is typically referred to as the “where” pathway.
As you can probably guess, the source of the information in the dorsal stream begins with the rods in the retina and the magnocellular cells in the LGN: these are tuned to fast changes rather than sustained details. This information works its way through V1 and V2 into area V5, an area spe- cialized for motion detection, and onward into the parietal lobe (FIGURE 5.23). As a result of the information in this stream, the posterior parietal cortex is necessary for understanding
FIGURE 5.23 The dorsal stream moves from V1 into the parietal lobe.
V1
W here
FIGURE 5.24 Motion can be seen even when there is no change in position. In the optical illusion shown here, leaves appear to wave.
FIGURE 5.22 Different representations of faces. The details between these faces trigger different responses in early visual areas. however, all are recognized as faces by neurons in the fusiform face area; the details are abstracted away.
05-Eagleman_Chap05.indd 146 02/11/15 3:33 pm
Higher Visual Areas 147
# 158305 Cust: OUP Au: Eagleman Pg. No. 147 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
(Exner, 1875). This demonstrated that the perception of motion does not depend on detecting positions and then comparing them.
But can the visual system work directly with motion sig- nals and not care much about position? In many cases, yes. Just using direct motion signals, animals can intercept, cap- ture, and avoid. But this seems strange, doesn’t it? If position is not always represented, how could a baseball outfielder possibly know where to go to catch a ball? The answer is that the outfielder does not need to use position—or velocity or acceleration—to catch the ball. W hen he runs to catch a fly ball, his brain (unconsciously) converts the temporal prob- lem (determining acceleration) into a spatial one by selecting a running path that keeps the optical trajectory of the ball going in a straight line (McBeath, Shaffer, & Kaiser, 1995). This way, spatial cues guide him toward the ball’s destination point, and acceleration never needs to be computed. As a side note, this strategy does not tell a person precisely where the ball will land, but instead only how to keep running—as illustrated by baseball players who crash into walls when chasing pop fly balls. This model of catching was verified by aerial photography, which quantified outfielders’ trajectories. Indeed, the paths they run in are not straight; they are curv y, as guided by the algorithm their brains employ (McBeath, et al., 1995). This strategy is also used by fighter pilots when pursuit tracking (O'Hare, 1999), fish ( Lanchester & Mark, 1975), and hoverflies (Collett & Land, 1975).
The bottom line is this: to catch or intercept a moving object, the visual system does not need an explicit representa- tion of position, or even velocity or acceleration. This counter- intuitive finding merely reinforces that we have little intuitive access to the mechanisms of the visual cortex that underlie our abilities (Eagleman, 2011).
appear to move if they happen to activate motion detectors in the right way.
Such motion illusions occur because the exact shading in the pictures happens to stimulate the visual system’s mecha- nisms for detecting motion. W hen those are activated, you see motion—whether or not something is actually moving out there (Mather et al., 1998). W hen the right keys are turned, your perceptual system has an experience as genuine as any other motion you’ve seen. We’ll turn to a remarkable illustration of this principle now.
This patient’s condition, known as motion blindness, highlights the strange fact that motion and position are sepa- rable to the brain. To a physicist, motion is change in posi- tion; to the brain, motion is painted on—as we see with illusions like waving leaves and rotating snakes.
How, then, do we sense motion? To a physicist, motion is change in position through time, but not so for the brain. From the point of view of the brain, motion detection can be achieved with neural mechanisms that are directly sensitive to a target’s velocity, without regard to position. The first ex- ample of this appeared in 1875, when the Austrian physiolo- gist Sigmund Exner (1846—1926) demonstrated that the detection of motion does not depend on position change. He rigged two electric sparks to appear in rapid succession, one beside the other. At the right distance in space and time, ob- servers perceived the motion of a single spark from the point occupied by the first to the point occupied by the second. Exner then moved the sparks closer together until the two sparks—when set off simultaneously—could not be distin- guished from a single spark. So although the two positions were now indistinguishable, a sequential sparking still caused the perception of motion. The direction of motion was distinguishable with a gap as small as 15 milliseconds
CASE STUDY: The World in Snapshots In 1978, a woman we’ll call Melissa was locked in her garage while her car was running. She sustained carbon monoxide poisoning. Fortu- nately, she lived, but unfortunately, the oxygen deprivation caused irre- versible damage to a specific region of her visual cortex: area V5, which is necessar y for the representa- tion of motion. The remainder of her visual system escaped injury, so she
could still see objects and their po- sitions with no problems (Zihl, von Cramon, Mai, & Schmid, 1991).
However, Melissa could no longer see motion (Zihl, et al., 1991). Stand- ing on a sidewalk, looking around to cross a street, she might see some- thing like a blue truck on the road. A moment later she would see that truck closer to her. Then, she would see it right in front of her. But
everything she perceived was in snapshots; the truck had no move- ment to it. As a result, she could no longer perform even the most basic tasks. Pouring water from a pitcher into a glass became a challenge. Rather than seeing the liquid move smoothly, it would seem to Melissa that she was catching glimpses of it through a strobe light.
05-Eagleman_Chap05.indd 147 02/11/15 3:33 pm
148 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 148 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Consider hemineglect, a bizarre disorder in which a person will disregard one half of the world. This is typically caused by brain damage (usually a stroke) to the right pari- etal lobe, which causes total neglect of everything on the left side of the person (Vallar, 1993). There is nothing wrong with a person’s visual system—instead, the problem is purely one of placing his attention any where in the left side of the world. Such patients behave as though one half of the world does not exist. They are completely unaware of the vanished half and do not usually miss it. For example, a typi- cal hemineglect patient will only dress the right side of his body, shave the right side of his face, and eat the dinner on the right side of his plate (Tsirlin, Dupierrix, Chokron, Coquillart, & Ohlmann, 2009). These bizarre behaviors are not caused by a blindness to the neglected side; instead, the problem is one of attention (in fact, the disorder is often called “hemi-inattention”). This can be demonstrated by calling a patient’s attention to his neglected side—say, by presenting a toy snake there or more simply by getting him to reach over into his neglected field with his opposite hand. Such tricks will help a subject notice the unattended side— demonstrating that he is not blind there—but this shift of at- tention does not last. Once the subject is distracted from the intervention, he will again neglect the left side (Tsirlin et al., 2009). We will return to hemineglect in detail in Chapter 8. For now, the important principle is that the dorsal stream steers the spotlight of attention.
This fact is made even more apparent in Balint’s syndrome, a disorder caused by damage to the parietal lobes on both sides (FIGURE 5.26a).
A lthough patients with Balint’s have a functioning ven- tral stream (and can therefore consciously recognize ob- jects), their loss of attentional steering steals away their ability to comprehend the big picture of a visual scene. Con- sider a patient with Balint’s, as he uses great effort to describe the scene in FIGURE 5.26b. He slowly reports what he sees: “. . . a car, a police car, another police car, a bus, a gorilla, a build- ing , a man (standing on a bus).” A lthough his ability to see details is unimpaired, the patient is unable to see the larger picture of the scene (a giant ape running amok through a city) because his attentional systems are not functioning correctly.
Part of the patient’s difficulty in understanding the scene stems from his simultagnosia, another symptom of Balint’s syndrome. Simultagnosia is an inability to recog- nize multiple elements in a scene (FIGURE 5.27) (Bose, 2008) and therefore the visual field as a whole. Imagine you placed a cell phone and a bottle of water in front of one another on a table. You then ask a patient with Balint’s syndrome to describe what he sees. He would tell you, “a cell phone.” A ny thing else? “No.” Then, a few moments later, he might say, “I see a bottle of water.” A ny thing else? “No, just the bottle of water.” He cannot see both at the same time. You can imagine the difficulty one would have, then, in describ- ing a complex scene: it would be impossible to see the forest for the trees.
Attention and the Dorsal Stream Take a look at FIGURE 5.25 and see how quickly you can name the number of horses in the scene.
Difficult, isn’t it? If the world were processed like a camera, you would have the answer instantly. But with our visual systems, it is impossible to process everything at once. Instead, to pull out the visual features of interest, you must select particular parts of the image for more detailed analy- sis. This process is known as focusing your attention.
The key to attention is that it is selective: it improves per- ception of stimuli that are attended to, and it interferes with the processing of those that are not. We will learn about attention in much greater detail in Chapter 8, but for now the key principle to appreciate is that attention can be spa- tial. At any moment, attention involves particular places in space. If you want to better understand what someone is saying, you can attend to her moving mouth. If you want to k now whether she’s happy or sad, you may watch the mus- cles around her eyes. If you hear her doorbell ring, you may shift your attention to the door way. In all cases, attention must be focused on an area. Like a spotlight, attention can be adjusted to span a larger or smaller area, but it cannot be split up into multiple areas. You may attend to small eye muscles or a large door way, but you cannot do both at the same time.
The dorsal stream is critical for guiding and adjusting the spotlight of attention. A lthough we do not usually think about the movements of our attention, its role becomes clear when there is damage to parietal lobe, a key component of the dorsal stream.
FIGURE 5.25 How many horses are in the picture?
05-Eagleman_Chap05.indd 148 02/11/15 3:33 pm
Higher Visual Areas 149
# 158305 Cust: OUP Au: Eagleman Pg. No. 149 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Lesions in the dorsal stream, on the other hand, result in problems with knowing how and where to interact with objects (Pisella et al., 2009). As we saw in the case study of the woman previously mentioned, damage to her motion- processing areas rendered her perception of the world as a series of snapshots. And patients who sustain damage to larger areas of their parietal lobes find themselves unable to move their attentional spotlight appropriately—as in the cases of hemineglect or Balint’s syndrome (Egly, Driver, & Rafal, 1994).
Let’s consolidate what we’ve just learned by thinking through two clinical cases. The first patient we’ll consider is named D.F. (Milner, 1995). Because of damage to her cortex, she is blind. She cannot name objects, and she cannot distin- guish among circles, triangles, and squares right in front of her. But strangely, she can grasp objects correctly. And if you ask her to put an envelope through a slot (one that can be tilted to any angle), she can do that perfectly well, although she claims to see neither the letter nor the slot (FIGURE 5.28).
Let’s compare D.F.’s case to that of another patient named A .T. He can see objects just fine, but he cannot pick them up them correctly (Goodale & Milner, 1992). W hen asked to pick up a small block of wood, he bluntly lowers his palm down toward it and tries to pull all his fingers together at once. A lthough he can see, he simply cannot make use of the visual information in the scene.
These two patients have opposite problems: the first can interact with objects correctly, but cannot see them. The second can see and recognize objects with no trouble, but cannot interact with them correctly.
W here are the areas of brain damage in these patients? At this point in the chapter, you should have no trouble
identifying that D.F.’s lesion is located in the ventral visual stream, which involves the temporal lobe, whereas A.T.’s is in the dorsal visual stream, which involves the parietal lobe.
Comparing the Ventral and Dorsal Processing Streams A good way to understand the differences between the ven- tral and dorsal streams is to compare what happens with damage to them. Bilateral damage to the face areas in the ventral visual stream, for example, will lead to a disorder known as prosopagnosia (face blindness) (Damasio, Damasio, & Van Hoesen, 1982). In this condition, the visual system functions, but one cannot recognize faces. W hen looking at the face made of vegetables in FIGURE 5.22, a person with prosopagnosia merely sees the pieces of produce.
FIGURE 5.27 A typical test for simultagnosia. A person is asked to look at a picture like this one and describe what he or she sees. A person with simultagnosia will report seeing only one object at a time.
FIGURE 5.26 Balint’s syndrome. Typical areas of damage leading to Balint’s syndrome. (a) When the dorsal stream is damaged in both hemispheres, a person can no longer use attention to understand the visual world. (b) When trying to describe a complex visual scene, a patient with Balint's syndrome will see the details, but not understand the bigger picture.
(b)
(a)
05-Eagleman_Chap05.indd 149 02/11/15 3:33 pm
150 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 150 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
appropriately with objects in the world, whereas damage to the latter can leave a patient unable to see or recognize objects. Keep in mind that both systems work together under normal circumstances. As Goodale and Milner remind us, “Both sys- tems have to work together in the production of purposive behavior—one system to select the goal object from the visual array, the other to carry out the required metrical computa- tions for the goal-directed action” (Goodale et al., 1994).
A lthough we have learned about the concept of a visual hierarchy, we will see in the coming sections that this story is not, in itself, complete. The brain is characterized by feed- back loops. As a result, although the ventral and dorsal streams are specialized, they interact with one another. As one example, we mentioned that patients with ventral stream lesions may have prosopagnosia, or face blindness. However, when shown a picture of a familiar face— compared to an unfamiliar face—a prosopagnosic will still have an increase in the galvanic skin response (a signature of the autonomic nervous system related to an emotionally significant event) (Fox, Iaria, & Barton, 2008). This suggests that face recogni- tion is partially accomplished in areas other than the classic “face areas” in the ventral stream. More generally, other, rare agnosias that are beyond the scope of this chapter reveal a
The Bigger Picture of the Visual Brain As we have been seeing, layers of visual processing are organized in a hierarchy. As signals move from the eyes to the LGN and through the visual cortex, processing changes from specific, signal-oriented responses to more abstract, object-oriented re- sponses. Neuronal populations become more specialized the higher in the hierarchy they live—from spots of light in retinal ganglion cells to face recognition in anterior temporal lobe neu- rons. At higher stages, neurons have larger receptive fields, and fewer neurons respond to any particular stimulus.
As a result of this hierarchical structure, damage to primary visual cortex leads to scotomas, areas of diminished vision or complete blindness in the visual field. Damage to secondary visual cortex leads to visual agnosias, loss of rec- ognition or meaning of objects. Damage to tertiary areas (such as those in inferotemporal cortex or parietal cortex) leads to specific deficits, such as face blindness or the inabil- ity to see motion.
We also learned in this section about the dorsal and ventral processing streams—the where and what pathways. Damage to the former can make a patient unable to attend to or interact
CASE STUDY: The Blind Woman Who Could See, Sort Of A patient we’ll call Amanda suffered a stroke that injured her primary visual cortex on one side. As a result, she now has blindness in one half of her visual field.
Imagine that you pick up a card- board shape, hold it up on Amanda’s blind side, and ask her to describe what she sees.
Not surprisingly, Amanda will report that she has no idea.
You encourage her to take a guess anyway. There’s obviously no consequence if she’s wrong.
But Amanda insists that she really is unable to tell you. She’s blind in that hemifield.
You assure her that’s fine, but that you still want her to guess.
Finally, eager to get you to quit asking, she floats a guess that the shape is a hexagon. To her surprise
and yours, her rate of guessing cor- rectly is well above chance (Weisk- rantz, 1990a,b). Some part of her brain is seeing unconsciously. This phenomenon is called blindsight. A recent demonstration of blindsight shocked research professionals: a doctor blinded by two strokes in his primary visual cortex was able to navigate an obstacle course in a hallway (de Gelder et al., 2008).
FIGURE 5.28 Patient D.F. (a) D.F. cannot correctly report the orientation of a slot. (b) If you ask D.F. to “mail” a letter through the slot at any orientation, she does so with no problem. (a) (b)
“I have no idea what the
orientation is…”
05-Eagleman_Chap05.indd 150 02/11/15 3:33 pm
Perception Is Active, Not Passive 151
# 158305 Cust: OUP Au: Eagleman Pg. No. 151 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
complexity of relationships among seeing, naming, recog- nizing, interacting, gesturing toward an object, pantomim- ing what to do with an object, and so on. Therefore, we should consider the separation of the ventral and dorsal streams a useful model that is clinically relevant, but not complete.
Consider another example of complexities beyond a simple visual hierarchy. We learned earlier that visual infor- mation follows a pathway from the retina to the LGN to V1. But that’s only true for 90% of the information leaving the eye via the retinal ganglion cells. The other 10% of informa- tion bypasses the LGN in favor of other destinations, includ- ing the superior colliculus, amygdala, and pulvinar nucleus of the thalamus. Do these subcortical pathways from the retina serve any purpose in vision? Indeed they do, as we can see from the following strange case.
These cases demonstrate that something in the brain is seeing—it’s just not the part of the pathway that depends on the integrity of the visual cortex. For many patients with blind- sight, the damage that makes them blind is within V1. How- ever, as we have described so far, seeing involves many other parts of the brain. We do not know how blindsight works, but the best-described theory involves the different pathways from the retina to the brain. The pathway from the retina to the LGN to V1 accounts for about 90% of the output from the eyes to the brain. The other 10% projects from the retina to other areas, including the superior colliculus and the pulvinar nucleus of the thalamus. These areas are important for visual attention, and it may be that, even with V1 damage, the subcortical path- ways like those of the superior colliculus and amygdala are still intact. These are sufficient to carry some degree of visual infor- mation, but, because of the damage to V1, that information does not rise to the level of visual awareness (Cowey, 2010).
Perception Is Active, Not Passive
Interrogating the Scene with Our Eyes In the late 1960s, the psychologist A lfred Yarbus set out to understand how the brain actively seeks information from the world—sometimes in ways we’re not aware of. To that end, he showed volunteers a copy of Ilya Repin’s painting An Unexpected Visitor (FIGURE 5.29). Their task was simply to answer questions, such as: “How wealthy are the people in the painting?”, “How old are they?”, “W hat were they doing just before he walked in?”, “How long had the visitor been away?” During the inquiries, Yarbus used an eye-tracker to measure where the participants’ eyes were looking.
We might assume that gazing upon a painting involves simply aiming your eyes there. However, with each question, the subjects’ eyes traversed the canvas in different patterns to ferret out the clues needed to answer the questions (Yarbus,
FIGURE 5.29 Eye movement recordings from a participant looking at The Unexpected Visitor. The different conditions were: (1) Free examination. Then, before the next recordings, participants were asked to (2) judge the wealth of the family, (3) estimate the ages of the people, (4) deduce what the family was doing just before the arrival of the unexpected visitor, (5) recall the clothes worn by the people, (6) remember the positions of the objects and people in the room, and (7) guess how long the unexpected visitor had been away. Each eye-recording lasted 3 minutes.
1967). When assessing the ages, they looked at the faces; when judging affluence, they focused on the possessions and clothes.
The brain directs the eyes to very specific locations to obtain the information it needs. It interrogates a scene to pick up details on a need-to-know basis. In other words, despite the impression that we see everything at once about a paint- ing, we don’t. And this is because the brain does not need to. It only requires a strategy that allows it to find where it needs to go when it comes time to answer a question.
A lthough you have lived with your pair of eyeballs from the beginning of your life, you have little direct knowledge about what they’re up to. Your introspection is limited; you can gain much more insight from watching other people’s eyes. To appreciate how little you know about your own eye movements, reflect on the rapid, precise saccades you’re making right now, as you read this page. This is just one dem- onstration of the fact that vision is not a passive process, but an active one.
05-Eagleman_Chap05.indd 151 02/11/15 3:33 pm
152 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 152 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
“filling in.” Your brain completes the scene, using informa- tion from around the missing region to fill the gap. As a result, you do not experience a hole (Ramachandran, 1992). In the case of FIGURE 5.30, you don’t see a chasm when the black dot disappears; you instead see a patch as though the page’s background has poured in. W hen the brain doesn’t have information from a region, it simply fills things in.
As with the earlier examples, the blind spot underscores the fact that you do not necessarily perceive what is out there. You perceive whatever your visual system tells you.
Seeing the Same Object Different Ways: Multistability As another example of the illusory nature of vision, con- sider a multistable percept. This is an ambiguous stimulus that can be perceived in more than one way and that typi- cally flips back and forth between the different options (FIGURE 5.31). You may have experienced the strange rever- sals of the Necker cube. A simple wireframe figure, it appears to you as a box oriented in a particular direction. However, if you continue to stare at it, it will reverse direction, appearing to come out of the page in a different way. The image remains exactly the same on the page, so whatever is changing must be inside your head. That is, your retina is receiving the same in- formation on its photoreceptors, but your brain is not just a passive recorder: instead it actively processes input to “see.” There is more than one way for the visual system to interpret the stimulus, and so it flips back and forth between the pos- sibilities (Bradley & Petry, 1977).
Binocular Rivalry: Different Images in the Two Eyes As we learned above, the brain exploits the slightly different information entering the two eyes to construct a perception of depth. But what happens when the information coming to the two eyes is not slightly different, but radically different? This situation leads to binocular rivalry (FIGURE 5.32). In bin- ocular rivalry, you don’t see both images simultaneously nor a fusion. Instead, you see one image, then the other, and then
The Blind Spot The brain does more than deduce what’s happening in the outside world—it often fabricates things entirely. For this reason, one basic feature of the retina was not discovered until 1668, many centuries after people could have easily no- ticed it. W hile studying a retina laid out in his laboratory, the French philosopher and mathematician Edmund Mariotte had a striking insight: where the retinal ganglion cell axons passed through the retina to become the optic nerve, there were no photoreceptors: there’s simply no room for them be- cause of the axons. W hy does this matter? Because if there are no photoreceptors in that region of the retina, it should be blind there. And it’s not.
Or isn’t it? Mariotte had to do some careful observation to determine something that no one had noticed before: in each eye, there is a spot in the visual field that captures no information—known commonly now as the "blind spot." Demonstrate it to yourself in FIGURE 5.30.
To experience the region where you have no vision, drag the page closer to and farther from your eyes. At some point, the dot seems to disappear. It’s now hitting your retina at the spot where you have no photoreceptors.
You might assume that you've never noticed your blind spot because it’s tiny, but in fact it’s huge. You can fit an area equal to 17 moons in your blind spot.
With so large an area of missing vision, why did it take so long for someone to notice before 1668? How did brilliant and curious minds like Michelangelo, Galileo, Newton, Shakespeare, and Kepler miss this basic observation? The reason it is so easy to miss is that we have two eyes, and their regions of blindness do not overlap. That means when both of your eyes are open, together they fully cover the scene. There is another and more significant reason as well: no one had noticed the blind spot because of a phenomenon called
FIGURE 5.31 Ambiguous figures. The Necker cube (a) can be perceived as coming out of the page in one of two ways, as can the cylinder (b). The staircase (c) can be viewed as going up and into the page or as the underside of a staircase coming out of the page. The perceived “switching” of the figure orientation takes place because “seeing” is an active process: nothing about the figure on the page has changed. (a) (b) (c)
FIGURE 5.30 The blind spot. Close your left eye and keep your right eye on the plus sign. Then, position the page closer to and farther from your face until the black dot disappears. At that distance, the dot is hidden in your blind spot.
05-Eagleman_Chap05.indd 152 02/11/15 3:33 pm
Perception Is Active, Not Passive 153
# 158305 Cust: OUP Au: Eagleman Pg. No. 153 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 5.32 Binocular rivalry. When two different images are presented individually to the two eyes, the brain does not perceive a fusion of both images. Instead, it perceives one image alternating sporadically with the other.
the first again (Blake & Logothetis, 2002). The visual system is locked in a battle between conflicting interpretations. You don’t see what’s actually there (two incompatible images); instead, you see a transitory perception of the image that’s winning at the moment. Nothing changes on the page, but your visual system is trapped in a dispute that tips back and forth. This parallels the multistable percepts in FIGURE 5.31, but here the rivaling interpretations are competing within the channels of the two eyes.
Binocular rivalry is not as simple as the information from the two eyes competing. Instead, the competition is between higher-level percepts. This can be demonstrated by swapping the two images presented to the eyes several times each second (such that each eye sees A, then B, then A . . .). This rapid alterna- tion of the images does not alter the amount of time it takes for the percept to change—usually a few seconds. In other words, a single phase of perceptual dominance can span multiple alter- nations of the stimuli (Logothetis, Leopold, & Sheinberg, 1996). This indicates that the neural representations of the two stimuli contend for visual awareness regardless of the eye through which they reach the higher visual areas. It is not the eyes that are competing, but the higher-level representations.
The examples of monocular rivalry (e.g., the Necker cube) and binocular rivalry (different images in each eye) lead to a question: why doesn’t a person perceive both op- tions at once? The fact that he does not suggests that the visual system forces a single outcome between competing percepts (Tong, Meng, & Blake, 2006). That is, perception appears to be a winner-take-all process. The strongest neural population at any moment determines the perception.
But if the system is really winner take all, then why does the percept switch instead of sticking to one conclusion?
One possibility is that although different populations of neu- rons compete for control of the percept, extended amounts of activity in one population cause it to fatigue, giving com- peting populations a chance to win (Orbach, Ehrlich, & Heath, 1963). A lthough fatigue may result from low-level functions (say, depletion of neurotransmitter vesicles), it can serve a higher-level adaptive strategy: if the percept is am- biguous, do not stay on one interpretation too long in case the other one may be more useful (Orbach et al., 1963).
Multistable figures and binocular rivalry both illustrate the active nature of perception. Nothing in the visual scene has changed, but your perception actively projects an inter- pretation onto the input.
We Don’t See Most of What hits Our Eyes: Fetching Information on a Need-to-Know Basis For many years, neuroscientists struggled to determine how the brain constructs a three-dimensional model of the out- side world. Eventually, it became clear that there was an as- sumption built into this pursuit, and a new idea arose: that the brain’s model is closer to a two and a half-dimensional sketch (Marr, 1982). This concept simply means that you don’t need to store a high-resolution model of the world out there—instead, you simply need to know the gist of it so you can determine where to look next. For example, you don’t need to actually store a complete representation of your bath- room counter. You have a general sense that there’s a surface
05-Eagleman_Chap05.indd 153 02/11/15 3:33 pm
154 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 154 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
There’s a strange consequence to this. How could it be that we’re missing so much information from a scene, but not at all aware that we’re missing that information? We turn to this now.
Vision Relies on Expectations We are not aware that we’re missing information from the visual world because what we see is not the raw data from the outside, but instead our internal model of the world. Let’s explore this concept with a few examples.
Change Blindness Two images are shown in FIGURE 5.33. Can you see the differ- ence between them? They look identical, but they’re not—and it will require some work for you to figure out where. Even if you overlapped and alternated the images on a screen, it would still be difficult to spot the difference. We’re surprisingly poor observers when it comes to discerning what’s changed be- tween two images, even if the changes are quite large. For example, there might be a crate in one image that’s not in the other, or a car, or an airplane—and those differences go unde- tected. To discern the difference between the images, we have to trawl the scene carefully. We have to analyze landmarks one at a time, comparing each sequentially. Once we find a mis- match, the difference then seems obvious—but it wasn’t at first. The lesson from change blindness is that we are not directly analyzing the visual input in front of us, but rather the rough sketch—or internal model—of what we believe is out there (Blackmore, Brelstaff, Nelson, & Troscianko, 1995; Rensink, O'Regan, & Clark, 1997; Simon, de Araujo, Gutierrez, & Nicolelis, 2006). It’s only when our model becomes updated that we see the next level of detail. (If you haven’t spotted it yet, the change in FIGURE 5.33 is the presence or absence of the
with several scattered items, and a mirror, and a wall. It’s only when you need the toothpaste that it becomes time for your visual system to search for it. If you were to wonder how much toothpaste is left in the tube, or what color the tube is, you’d be able to place your attention on the details, incorpo- rating the answer into your internal model. Even though the toothpaste tube was impinging upon your retinas the whole time you stood at the counter, you weren’t attending to its de- tails. You needed to fill in the finer points of the picture.
In the same vein, it’s often the case that we know some features of an object while we’re unaware of others. Imagine we asked you to look at these symbols and tell us what you saw: //////////////. You would easily describe that it is com- posed of diagonal lines. But if we now asked how many lines, you would find that difficult to answer. You can see there are a finite number, but you can’t quantif y them without some work. Thus, you can be aware of some aspects of a scene and not others; you become aware of further details (and realize what you’re missing) only when you ask yourself further questions.
The secret to understanding this aspect of the visual system is that an organism does not require all of the data in the world around it. It simply figures out how to retrieve the data it needs. It gathers information on a need-to-know basis. And what it needs depends on its goal at the moment.
If this view of vision comes as something of a surprise, just consider that you’re not aware of much at all until you ask yourself a question. W hat is the position of your tongue in your mouth? W hat does your right shoe feel like on your foot? W hat sound is the traffic outside making? Our typical state is to be unaware of most of the data hitting our sensory recep- tors. It’s only by turning our attention to small questions that we add those bits to our model. (We will explore attention in depth in Chapter 8). Prior to asking questions of the scene, we are usually unaware that we’re not seeing the whole pic- ture: we’re blind to our blindness. We typically walk around with a false confidence that we’re engaged in a complete pic- ture of the world. Despite the persuasive force of the illusion, however, we see mostly what our brains need to know.
FIGURE 5.33 Two slightly different pictures to demonstrate change blindness.
05-Eagleman_Chap05.indd 154 02/11/15 3:34 pm
Vision Relies on Expectations 155
# 158305 Cust: OUP Au: Eagleman Pg. No. 155 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 5.34 Visual illusion. The word “of” is repeated at the end of line 2 and the beginning of line 3.
The illusion of
of “seeing”
FIGURE 5.35 Light sources are preferentially seen as coming from the top. (a) Generally interpreted as bumps with a dimple in the middle. (b) Merely the same picture upside down—but it is now interpreted as dimples with a bump in the middle. (c) The same picture rotated 90 degrees; the depth information now appears ambiguous.(a) (b) (c)
black rectangle on the white cylinder at the left-hand side of the photo.)
As another example, imagine watching a short video of a person frying eggs. Now imagine the camera angle cuts to a different view. You would surely notice if the cook were sud- denly being played by a different actor, right? Strangely, the majority of observers fail to notice the change (Levin, 1997).
Consider another surprising demonstration of change blindness. In an experiment in Boston, random pedestri- ans were asked for directions by an experimenter. A s they were explaining a route, several rude work men carr y ing a door wedged themselves bet ween the t wo people. The con- versation resumed seconds after the work men passed . . . and most of the subjects did not notice that the person they were talk ing w ith (the experimenter) was a different person entirely, hav ing been swapped out w ith a person who had been hiding behind the door (Simons & Lev in, 1998). How did they miss it? Because they were only incor- porating a small bit of the outside world into their internal models.
Looking directly at something is no guarantee of “seeing” it. However, neuroscientists weren’t the ones to figure this out—magicians were. Their profession capitalizes on this knowledge (Macknik et al., 2008). Once a magician directs your attention, he or she can then execute sleights of hand in full view. Even though your retina is being struck by the pho- tons that convey the trickery, the magician can act with con- fidence that your brain won’t see it. You’re only going to see your expectations.
Our poor ability to see what’s really happening out there explains many events on our roadways, such as the number of traffic accidents that happen in plain view of pedestrians, other cars, or trains. In so many of these cases, a driver’s eyes are aimed in the right direction, but the brain isn’t seeing. Vision requires more than looking at the right place at the right time. In fact, we’re guessing you didn’t notice that the word “of ” appears twice in the triangle (FIGURE 5.34).
Saving Resources by Embedding Prior Experience Internal models of the outside world allow the brain to save time and resources and to come to rapid perceptual conclu- sions. Consider how the pattern of gradient-filled circles in FIGURE 5.35 produces a perception of three-dimensional bumps and dimples. Look at FIGURE 5.35a and note that the shading on the circles could be consistent with lighting from above or below—the information in the picture is ambigu- ous. However, if you are like most people, your brain assumes the light source comes from above—in keeping with the sun, moon, overhead lights, etc. (Ramachandran, 1988) This makes the middle circle in FIGURE 5.35a appear as a dimple while the surrounding ones appear as bumps. In FIGURE 5.35b we’ve simply turned the image upside-down. This change causes the bumps to become dimples and the dimple to become a bump, in keeping with a light source coming from its usual location: overhead. W hen the picture is rotated by 90 degrees (FIGURE 5.35c), the light source can be easily per- ceived as coming from either the right or the left; with a little practice one can switch the perception (bumps become dim- ples and vice versa), although that is much more difficult in FIGURE 5.35a and FIGURE 5.35b.
This assumption of overhead light sources allows the brain to save on computational resources, at least until fur- ther information demands it to reconsider. Since the lighting of a visual scene in our world is almost always from above, that provides a reasonable starting point for visual analysis.
Importantly, the visual system embeds its prior experi- ence with the world into our present perceptions. In this way, the system can opt for the most likely interpretation, based on what it’s seen before.
05-Eagleman_Chap05.indd 155 02/11/15 3:34 pm
156 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 156 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
likely one. This exemplifies how the brain employs prior ex- perience to settle on an interpretation of the highest likeli- hood from many possibilities.
Unconscious inference can be formalized through the mathematics of the English reverend Thomas Bayes (1702– 1761). A lthough the details of “Bayesian inference” are beyond the scope of this chapter, the concept is simple enough: the probability that a stimulus in the world is, say, possibility A or possibility B depends on the probability of each of those options (given the input on the retina), as well as the overall likelihood of each of those possibilities hap- pening in the world (Uka & DeAngelis, 2004). Let’s take the example of the illusion of movement based on the position of a shadow, described above. A lthough the moving shadow gives equal probability to either (A) a change in depth or (B) a moving light source, the likelihood of B is low in the face of previous experience (light sources don’t typically change position rapidly); thus interpretation A wins. If you lived on an alien planet on which the overhead light sources danced around frequently but objects rarely changed heights, you would interpret the stimulus in the opposite manner. Same input, different output based on prior experience.
To summarize where we are, we’ve seen that the visual system relies on internal models built from best guesses, given retinal input and prior experience. The important lesson is that the brain’s perceptions are not constructed from scratch, with no prior information. That was a mistake made by some early pioneers of computer vision, who as- sumed that visual systems have to parse all of the input and build visual objects from scratch (Knutsson & Granlund, 1994). In fact, vision involves the comparison of incoming information with detailed internal models of the world.
But how does the concept of an internal model mesh with what we learned of the visual system in the first half of the chapter? Didn’t it seem as though neural signals from the retina simply worked their way up the hierarchy until they were perceived? We’ll now see why that process is only half the story.
In another example of this point, the movement of a shadow is perceived as movement in depth (Mamassian & Kersten, 1996). Note that such “movement” is ambiguous: a shadow can move because of a change in the depth of an object or instead because the position of the light source has changed. In general, however, the visual system not only as- sumes the overhead source, but also a source fixed in position. In other words, the brain makes a reasonable assumption based on accumulated experience.
Unconscious Inference As we have learned in the previous sections, data from the world are not simply analyzed with unbiased interpretations. Instead, the visual system capitalizes on prior expectations.
A lthough this idea might be a bit surprising, it is not new. The German physician and physicist Hermann von Helm- holtz (1821–1894) was one of the first people to entertain this model of perception. He suspected that the small amounts of information dribbling in through the eyes were too slight to account for the rich experience of vision. He therefore deduced that the brain makes assumptions about the incoming data, based on previous experiences (von Helmholtz, 1867). By this method, the brain can use its best guesses to rapidly turn a little information into a much larger picture.
In this sense the brain is a hypothesis generator, building models of the world and striving to verify them based on input through the tiny windows of the senses. As a result, we often see what we are expecting to see and hear what we are expecting to hear.
Helmholtz thus proposed that vision arises from a pro- cess of “unconscious inference.” The word inference means the brain infers, or deduces, what is likely to be out there, and unconscious simply reminds us that we’re unaware of the de- ductions. We have no direct access to the neural machinery that analyzes the statistics of the world. We simply perceive the end result of the calculations.
As an example of this point, let’s return to depth percep- tion. Earlier we learned that the visual system can judge depth from the disparity between the two eyes. However, this is only useful out to 30 meters—past that distance, the line of sight to distant objects falls on nearly corresponding points of the retinas (Coltekin, 2009). As a result, the brain uses other cues to determine depth (Swan, Jones, Kolstad, Livingston, & Smallman, 2007). One of them comes down to mere experience with the world. The visual system builds up prior expectations about the relative sizes of objects (Swan et al., 2007). Even in the minimalist cartoon of FIGURE 5.36, you would probably guess that the cat is closer to you than the artist, who is closer than the Eiffel Tower he is painting in the distance. This doesn’t have to be the case: it could be a giant feline eyeing the artist he is about to eat, who in turn is painting a small model of the Eiffel Tower. However, the latter interpretation doesn’t seem the most
FIGURE 5.36 Our prior knowledge about size is a cue for depth. Because we know the size of cats, artists, and the Eiffel Tower, we assume that they exist in that order of nearness to us in the cartoon. An alien from a planet with giant cats and tiny towers would interpret the depth in the scene differently.
05-Eagleman_Chap05.indd 156 02/11/15 3:34 pm
Vision Relies on Expectations 157
# 158305 Cust: OUP Au: Eagleman Pg. No. 157 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
line. In fact, it is such an easy mistake that almost all early theories (and many unsophisticated modern theories) begin with this assumption. However, there is a critical clue that the assembly-line model does not account for. A long with brain wiring that proceeds in the forward direction (say, from hypothetical areas 1 to 2 to 3), there are feedback loops (from 3 to 2 to 1). In fact, there is just as much feed- back in the brain as there is feedforward, a concept referred to as recurrence or loopiness (Bell 1999). Once theoreti- cians began to consider the feedback seriously, instead of sweeping it under the rug, this opened up brand-new models of brain function. For example, one model suggests that the visual system should not be understood as a succession of signals climbing a hierarchy; instead, it should be under- stood as a reverse hierarchy: a system that makes a best guess about the essence of a scene (“there is an animal moving my way”) and then sends that guess down to succes- sively lower areas to see if the incoming details are consistent (A hissar & Hochstein, 2004).
Such a model is consistent with the fact that a sufficient amount of feedback allows a system to run “ backward.” Try closing your eyes and imagining a scene—rays of sun break- ing through clouds over a snow y mountaintop, or a banana sundae festooned with chocolate sprinkles. W hen you imag- ine those scenes, your visual cortex becomes active. You’re not actually gazing upon those items—but when you visual- ize them, the higher regions in your visual cortex send infor- mation down to lower areas. Obviously, under daily circumstances your eyes are involved in pushing signals up to your higher areas—but connections going the other way make it possible for the cortical areas to get by without any contributions from the eyes at all. W hen you visualize, you run the whole system in reverse.
As it turns out, recurrent networks have a few great ad- vantages over a simple assembly line model. For one thing, a system that can generate internal events can build up repre- sentations of the world—enabling it to outperform anything a feedforward system could do. Consider the act of trying to hit a rapidly flying racquetball. If your visual system were only feedforward, you’d always be swinging for the ball at some distance behind its actual position. W hy? Because signal processing takes time. But now consider a recurrent system that has a representation of Newtonian physics: it can make a guess about exactly where the ball is right now and where it is going (Lacquaniti & Carrozzo, 1993; Tresilian, 1999; Zago et al., 2004). A feedforward system is always play- ing catch-up, but a system that deals with an internal model can incorporate effects of gravity to estimate when and where the ball will make contact with the racquet (Wolpert & Fla- nagan, 2001; Wolpert & Miall, 1996). Using this trick, the brain can make predictions rather than being bound to only the latest sensory data that comes in.
This is a specific example of the broader concept of inter- nal models of the world. Beyond predicting the physics of moving objects—so that you can intercept or dodge— internal models may also be the key to understanding our
Activity from Within In the traditional model of perception, information from sensory receptors streams along pathways into the brain, eventually becoming sight, hearing, touch, taste, and so on when they reach an end point. However, that model covers only part of the story. In reality, much of the brain’s activity comes from within (Llinás, 2002; Eagleman, 2011).
If that seems strange, consider other examples of internally generated activity: all the details of your breathing and diges- tion are controlled by internal activity in your brainstem. As it turns out, this happens not only in the brainstem, but through- out the rest of the brain as well. Most neural activity is pro- duced on the inside, and it is merely altered by external sensory input. Consider what this means: the visual experiences you have when you’re dreaming are really the same thing as the visual experiences you have while you’re awake (Llinás, 2002). The only difference is that in your waking life, your eyes are open, allowing a bit of activity from the outside world to anchor your experience; during dreaming, on the other hand, the visual system is free to roam. Minds roam in other situations as well. For example, people in sensory deprivation chambers and prisoners in solitary confinement both experience hallucina- tions (Vernon, Marton, & Peterson, 1961; Vernon, McGill, & Schiffman, 1958). As the external input wanes, the internal ac- tivity picks up the slack (Allan, 1977; Eagleman, 2011).
As an example of the interplay of internal activity and exter- nal data, consider a disorder known as Charles Bonnet syndrome. Characterized by visual hallucinations, this syn- drome affects ten percent of patients who have visual loss due to eye disease (Gillig & Sanders, 2009). As they lose their sight, they begin to see things—such as flowers, birds, people, buildings— that are not real. Bonnet, an eighteenth-century Swiss philosopher, first described this phenomenon after ob- serving his grandfather (Berrios & Brook, 1982). He noticed that as his grandfather lost his vision to cataracts, he would at- tempt to interact with things and people who weren’t there. As the input of external data slows down, the internal activity is enough to produce an outside world. The syndrome is essentially equivalent to an intrusion of dreaming into the waking state.
The surprising bottom line is that normal vision is hardly different from hallucinations; they differ only by the degree to which they’re anchored by external data. Hallucinations are normal vision untethered.
Together, these observations present a more realistic model of the visual system. Let’s now circle back to the anat- omy to understand how the internal activity and external data interact.
Feedback Allows an Internal Model It’s a common mistake to think of the brain like a computer— or an assembly line—with signals coming in, getting pro- cessed in successive stages, and finally reaching some finish
05-Eagleman_Chap05.indd 157 02/11/15 3:34 pm
158 PART 2 • ChAPTER 5 Vision
# 158305 Cust: OUP Au: Eagleman Pg. No. 158 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
LGN than axons going the other way. This is just what we’d expect from a system that sends detailed predictions “down” to the LGN, and only returns the difference signal back “up.”
Once we begin to think about the brain as a heavily re- current network that generates its own models about the world, other mysteries begin to resolve. Consider the disor- der known as Anton’s syndrome, in which a person be- comes blind (usually from a stroke) but denies his blindness (Symonds & Mackenzie, 1957). An interaction with a patient with Anton’s syndrome might go something like this: you hold up two fingers and ask the patient to report how many fingers. Without hesitation, the patient says, “five.” You ask what color your shirt is, and the patient confidently gives an answer, even though the response has no relation to the color you’re actually wearing.
It’s not that the Anton’s patient is lying (or embarrassed or roguish)—it’s simply that he’s not experiencing blind- ness. He’s seeing, because his internal model is still cranking along—even though it’s not meaningfully anchored to the outside world. It often turns out that patients with Anton’s do not go to the doctor’s for a while after they experience a stroke—because although they are blind, they don’t know that they are blind. They begin to realize something strange is happening only after repeated collisions with furniture and people. To summarize, the Anton’s patient makes strange claims about the number of your fingers or the color of your shirt because he is seeing his internal model—a model that’s normally anchored by external data, but has now become unmoored from reality. If it seems difficult to imagine what this would be like, just consider that you have this experience every night when you dream.
And this leads us to highlight an important distinction between the words sensation and perception. Sensation is the detection of a signal. For example, when an electric motion detector turns on your garage lights, we say your car’s movement has been sensed. But we have no reason to think that the motion detector, a simple circuit, perceives—it is perfectly adequate to assume that electrons move through the wires of the detector, pushing the current above a thresh- old voltage to activate the lights. The first sections of this chapter taught what’s known about the sensation of photons at the retina and the passage of signals from there. But for the perception of the world, much more is required—specifically, an internal model against which the input is constantly com- pared for mismatches.
Conclusion Attempts to build artificial computer vision used to assume that the human visual system produces a three-dimensional, high- resolution model of the world, distinguishing and representing all objects. That turned out to be a misdirected approach and an impossible feat to realize. Scientists now appreciate that they do not need to solve the complete-model-of-the-world problem:
conscious awareness. Specifically, some frameworks sug- gest that your view of the world is not assembled from in- coming data (the way a camera would “see” the world), but instead perception only arises when your expectations (“it’s an animal”) successfully match the incoming data (e.g., Grossberg, 1980).
As strange as this seems, consider it in light of something you already know: that your expectations influence what you see. Try to make sense of FIGURE 5.37. Difficult, right? Your brain doesn’t have an expectation about this pattern, and so you see nothing but…blobs. Without a way to con- nect your predictions and the incoming photons, you “see” very little.
Given all of these considerations, let’s take a fresh look at the visual system. In the model we’ve been building up to, the job of the visual cortex is to construct expectations of what the world “out there” will look like. In the 1950s, neuro- scientist Donald MacKay proposed that the visual cortex builds an internal model—the purpose of which is to predict the information that comes up from the eyes to the cortex (MacKay, 1956). With this as the goal, the visual system uses feedback connections to tell the LGN what it anticipates. The LGN then uses feedforward connections to report on the difference between what actually came in and what was predicted. W hen the visual cortex receives the information about that difference—known as the prediction error—it then updates its internal model. The aim is to have less error next time—that is, to learn from its mistakes. This model is consistent with a strange fact about the anatomy: there are ten times more axons projecting from the visual cortex to the
FIGURE 5.37 The role of expectation in perception. Perception requires a model of what is being seen. Although a figure is in these blobs, it is difficult to see. Only after receiving a hint (see Figure 5.38) does the figure appear as something interpretable. Figure from Ahissar and hochstein (2004).
05-Eagleman_Chap05.indd 158 02/11/15 3:34 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 159 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Key Principles 159
impinging on his retinas, becoming transduced into signals that traveled to his LGN and on to his visual cortex.
But when the bandages were removed from his eyes and his children walked in the room, he couldn’t see them. His brain didn’t know how to interpret the torrent of action po- tentials coursing up the optic nerves. As you now know, this is because Mike’s brain did not know how to make an internal model of the world to be compared to the new external data. The incoming signals were not sufficient by themselves—his visual system needed expectations against which to compare them. Without that, nothing could be perceived.
By practicing for several weeks—reaching for objects, touching them, knocking on them—Mike was able to align his expectations with the incoming signals. And then there was light.
P.S. If you are still struggling with FIGURE 5.37, see the hint in FIGURE 5.38.
not everything in the outside world needs to be represented. In- stead, the information needed depends on the goal: it is task dependent. When you want to find the correct gate at the air- port, you do not need to process and represent all the light sig- nals hitting your retina. You do not need to represent the details of other passenger’s faces, the titles of the books in the gift shop, the color of the carpet, the shape of the ceiling tiles, and the signs displayed at other gates. You simply need the information that will get you to your gate and allow you to recognize it. That information will differ if you are looking for a snack or a rest- room. The selection of information happens via attention, which we will return to in Chapter 8.
We have also seen that perception is the result of infer- ences about what is probable, based on prior experiences. These unconscious inferences result in rapid perceptions of the most likely scenario given the data. Thus vision is deter- mined by two streams: the ongoing input of external data and prior expectations. Note that expectations can be both genetic (nature) and experiential (nurture)—take as an ex- ample a horse bucking at the sight of a snake (genetic) and salivating at the shape of the metal feeding trough (experien- tial). The main challenge for a brain is to compute through many alternative interpretations of the input and to settle on the best one (or at least the best one for the moment). In this way, perception is an active process.
The first half of this chapter presented the anatomical framework of a hierarchy in the visual system—one in which signals move from the retina to the LGN to the sprawling regions of the visual cortex. By the end of the chapter, we ap- preciated that the concept of a hierarchy, although useful to understand the anatomy, is not complete. There is no “finish line” where signals reach perception. Instead, like all parts of the brain, the visual system is a massively recurrent (loopy) system. In the end, the most accurate way to understand the visual system is to recognize its confluence of forward and reverse directions—in other words, internal expectations meeting up with external data.
Let’s now return to Mike May, the blind downhill skier whom we met at the beginning of the chapter. After the opera- tion that cleared his corneas, it seemed as though his vision would be functional: after all, photons were now successfully
FIGURE 5.38 With a simple hint, the blobs in Figure 5.37 instantly take on meaning as a bearded man. Now that you’ve established an expectation, you can’t “un-see” the bearded man in Figure 5.37.
KEY PRINCIPLES
• Although seeing the world seems effortless, approximately 30% of the brain is devoted to constructing vision.
• Photoreceptors in the retina transduce photons into neural signals that move into higher parts of the visual system.
• The visual system is hierarchical, building from fine details to larger concepts.
• Damage to lower parts of the visual system leads to a lack of sensation (that is, blindness), whereas damage to successively higher areas leads to more specific deficits of perception (e.g., ability to see but inability to recognize).
05-Eagleman_Chap05.indd 159 02/11/15 3:34 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 160 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
160 PART 2 • ChAPTER 5 Vision
• The visual scene relies on internally generated activity as much as on data from the outside world.
• The visual system contains both feedforward and feedback projections, making it “loopy.”
• Much of what we see comes from unconscious inference—that is, our expectations of what we believe is “out there.”
• As visual information becomes more abstracted at higher levels of the visual system, the information moves in two distinct processing “streams”: the ventral stream and the dorsal stream. The ventral stream deciphers what objects are—in other words, how to identify and categorize them. The dorsal stream is focused on where objects are and how to interact with them.
• Vision is active, not passive. Without being con- sciously aware of it, we actively interrogate the world with our eyes, pulling details into our internal models.
KEY TERMS
Visual Perception Mach bands (p. 132) sensory transduction (p. 133) hair cells (p. 134)
Anatomy of the Visual System cornea (p. 134) iris (p. 134) pupil (p. 134) lens (p. 134) retina (p. 134) retinal ganglion cells (p. 134) amacrine cells (p. 134) horizontal cells (p. 134) phototransduction (p. 134) rods (p. 135) cones (p. 135) fovea (p. 135) receptive field (p. 136) center-surround structure
(p. 136) contrast enhancement (p. 136) optic nerve (p. 138)
blind spot (p. 138) optic chiasm (p. 138) parvocellular retinal ganglion
cells (p. 139) magnocellular retinal
ganglion cells (p. 139) optic radiations (p. 139) striate cortex (p. 139) simple cells (p. 139) complex cells (p. 139) blobs (p. 140)
Higher Visual Areas secondary visual cortex (V2)
(p. 142) tertiary visual cortex (p. 142) ventral stream (p. 143) dorsal stream (p. 143) inferotemporal cortex (p. 143) position invariance/size
invariance (p. 144) sparse coding (p. 144) all-or-none (p. 144) fusiform face area (p. 144)
area V5 (p. 146) motion blindness (p. 147) attention (p. 148) hemineglect (p. 148) Balint’s syndrome (p. 148) simultagnosia (p. 148) prosopagnosia (face
blindness) (p. 148) blindsight (p. 150)
Perception Is Active, Not Passive multistable percept (p. 152) binocular rivalry (p. 152)
Vision Relies on Expectations internal model (p. 154) change blindness (p. 154) Charles Bonnet syndrome (p. 157) feedback loops (p. 157) recurrence (p.157) reverse hierarchy (p. 157) Anton’s syndrome (p. 158) sensation (p. 158) perception (p. 158)
REVIEW QUESTIONS
1. Explain how the responses of a simple cell in V1 can be constructed from the input it receives from several LGN cells. How can a complex cell’s responses be constructed from the input it receives from several simple cells?
2. Explain the difference between the dorsal and ventral visual processing streams.
05-Eagleman_Chap05.indd 160 02/11/15 3:34 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 161 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Critical-Thinking Questions 161
7. How are blindsight and Anton’s syndrome like opposites of one another?
8. Scientists have discovered that there is as much feedback as feedforward circuitry in the brain. How does this challenge the assembly-line model of visual processing?
9. Give an example in which prior experience influ- ences what is perceived.
10. Define sensation and perception. Why does there need to be a distinction drawn between these two terms? Can one exist without the other?
3. What might be the expected result of damage to the eyes? To the primary visual cortex? To the dorsal stream? To the ventral stream?
4. Would we still have depth perception if our eyes were one above the other instead of side by side? Why or why not?
5. In binocular rivalry, why does the perception switch back and forth instead of sticking to only one interpretation?
6. Why do prisoners who have been put in pitch-black solitary confinement cells begin to hallucinate?
CRITICAL-THINKING QUESTIONS
1. From an evolutionary perspective, why do you think that the human brain has separate dorsal and ventral processing streams? Why would it not have just one processing stream that served the functions of both the dorsal stream and the ventral stream?
2. Do you think that some areas of the human brain that are devoted to vision in sighted individuals are devoted to other functions in blind individu- als? What research methods would you use to test your hypothesis? Explain your reasoning.
3. Imagine that Mike May, the blind downhill skier whose case study you reviewed at the beginning of the chapter, had been 6 years old at the time of his successful corneal transplant (and resto- ration of sight) instead of 46 years old. Would you expect his visual perception to be different in that case? Explain. Would you expect to see a different time course, with therapy, for the im- provement of his visual perception? Explain your reasoning.
05-Eagleman_Chap05.indd 161 02/11/15 3:34 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 162 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
162
LEARNING OBJECTIVES By the end of this chapter, you should be able to: • Explain how the principles of perception and
transduction that we learned about in vision (Chapter 5) also apply to hearing, touch, smell, and taste.
• Describe how sound is transduced from pressure waves in the air to electrical signals in the nervous system.
• Characterize the importance of labeled lines for the transduction of sensory signals.
• Differentiate the five major types of information perceived by the somatosensory system.
• Compare and contrast the chemical senses and the other senses.
• Provide two examples of phenomena that support the idea that the brain integrates information from the different senses.
• Explain why time perception could be considered a sense and the difficulties involved in studying it.
06-Eagleman_Chap06.indd 162 04/11/15 4:33 pm
# 158305 Cust: OUP Au: Eagleman Pg. No. 163 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
163
Other Senses STARTING OUT: The Man with the Bionic Ear
Detecting Data from the World Hearing
RESEARCH METHODS: Psychophysics
NEUROSCIENCE OF EVERYDAY LIFE: The Undetectable Cell Phone
The Somatosensory System
CASE STUDY: The Pain of a Painless Existence
Chemical Senses The Brain Is Multisensory
CASE STUDY: The Paralyzed Supreme Court Justice Who Claimed He Could Play Football
Time Perception
CHAPTER 6
06-Eagleman_Chap06.indd 163 04/11/15 4:33 pm
164 PART 2 • ChAPTER 6 Other Senses
# 158305 Cust: OUP Au: Eagleman Pg. No. 164 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
STARTING OUT The Man with the Bionic Ear
Michael Chorost (FIGURE 6.1) was born mostly—but not entirely—deaf because of a case of rubella during his mother’s pregnancy. Michael functioned just fine during his young adult life with the help of a hearing aid. But one day, while wait- ing to pick up a rental car, the bat- tery to his hearing aid died. Or so he thought. He replaced the battery but found that all sound was still missing from his world. He drove himself to the nearest emergency room and discovered that the re- mainder of his hearing—his thin au- ditory lifeline to the rest of the world—was gone for good. For un- known reasons, his only functioning ear had suffered “sudden-onset deafness.”
A hearing aid would do him no good now. Hearing aids work by blasting a partially functioning ear with increased volume. In other words, a hearing aid takes sounds
from the world and re- broadcasts them more loudly into the ailing au- ditory system. But this strategy only works if ev- erything beyond the ear- drum is functioning. If the inner ear is defunct, no amount of volume amplifi- cation will solve the prob- lem. Given the available technology, it seemed as though Michael’s enjoy- ment of the world’s sound- scapes had come to a sudden end.
But then he found out about a single remaining
possibility. In 2001, Michael under- went surgery for a cochlear implant. This tiny device circumvents the broken hardware of the inner ear to speak directly to the functioning nerve bundle just beyond it. It is es- sentially a minicomputer implanted directly into his inner ear, one that receives sounds from the outside world and passes them to the audi- tory nerve by means of 16 tiny elec- trodes (Wilson et al., 1991).
Although the damaged part of the inner ear is bypassed, the ex- perience of hearing doesn’t come for free. Michael had to learn to in- terpret the foreign language of the electrical signals being fed to his auditory system. Here is Michael’s description of his experience:
When the device was turned on a month after surgery, the first sentence I heard sounded like “Zzzzzz szz szvizzz ur brfzzzzzz?”
My brain gradually learned how to interpret the alien signal. Before long, “Zzzzzz szz szvizzz ur brfzzzzzz?” became “What did you have for breakfast?” After months of practice, I could use the telephone again, even con- verse in loud bars and cafeterias. (Chorost, 2011)
Although a minicomputer im- plant sounds like science fiction, cochlear implants have been on the market since 1982. Tens of thousands of people are walking around with these bionics in their heads, according to the National Institute on Deafness and Other Communication Disorders (2011). The software on the cochlear im- plant is hackable and updateable, so Michael has spent years im- proving the quality of the implant without further surgeries. Almost a year after the implant was acti- vated, he upgraded to a program that gave him twice the resolution. And it will only improve. As Michael puts it, “While my friends’ ears will inevitably decline with age, mine will only get better” (Chorost, 2011).
How does the cochlear implant work? And what does it teach us about the brain and its thirst for in- formation from the outside world— not just via hearing but by all the senses? And how do the other senses utilize the same principles we’ve already learned about in regard to vision in Chapter 5? In this chapter, we’ll find out.
FIGURE 6.1 Michael Chorost. Michael was born mostly deaf because of a case of rubella during his mother’s pregnancy but regained his hearing after undergoing surgery for a cochlear implant.
06-Eagleman_Chap06.indd 164 04/11/15 4:34 pm
Hearing 165
# 158305 Cust: OUP Au: Eagleman Pg. No. 165 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
Detecting Data from the World In the previous chapter we learned about vision as an inroad to several general principles of perception. Remember that trans- duction is the process of converting energy or information outside the body into the neural code used inside the body. In this chapter, we will extend the principles of perception and transduction to hearing, touch, smell, and taste. After all, you may have begun reading the previous chapter with the impres- sion that vision was effortless and, with minor exceptions, accurate—in other words, that your eyes were essentially like a high-resolution video camera. If that’s where you started, you learned an interesting set of lessons to the contrary. Now extend that reasoning. It may seem intuitive to think of your auditory system as two high-fidelity microphones. Likewise, it may seem reasonable to believe that your fingertips simply reach out and detect what is out there. As we’ve learned al- ready, our intuition often suggests ideas that are incorrect.
The person on the street will usually tell you that there are five senses: vision, hearing, touch, taste, and smell. But as we will see, there are many other ways of detecting data from the outside world, including vibration, pain, temperature, accel- eration, head position with respect to gravity, joint position and motion, pheromones, and the sense of time. There are also critical sensory signals that we detect from inside the body, such as stretching in our bladder or our gut. As we will see, not all senses are of equal importance to survival: although the visual system is highly represented in human brains, one can live without vision much more easily than one can live with- out, say, the sense of pain or the sense of body position.
W hat exactly is a sense? A sense is built from specialized cells that respond to particular physical phenomena, and the sense corresponds to brain networks that receive, interpret, and act on those signals.
A ll sensory systems fundamentally attempt to accom- plish the same goal: detect useful information sources from the world. W hereas your eyes are sensitive to electromag- netic radiation in a particular wavelength (as we saw in the previous chapter), your ears pick up information contained in air compression waves. Your fingertips sense temperature and pressure data, and your nose and mouth detect chemi- cals in the air and in objects.
Our relatives in the animal kingdom have developed an endless variety of exotic ways of picking up information from the world. The ghost knifefish of South A merica has specialized sensors to detect disturbances in the electrical field that surrounds it. In some species of snake, specialized detectors called heat pits allow the snake to detect energy in the infrared range. Many insects, birds, and even cows have small particles in their heads that allow them to orient to— and navigate by—the earth’s magnetic field. For our aquatic mammalian neighbors, whales and dolphins, the murkiness
of water makes it an environment that is ill suited for vision—and so they have developed sonar as an imaging strategy.
Regardless of the sensory system considered, the infor- mation is sent to what is known as the primary sensory cortex. A common feature across the senses is the concept of a map: information is laid out in specific ways on the surface of the brain. As we saw in Chapter 5, in the primary cortex of the visual system (V1), the map is retinotopic: neighboring neurons represent neighboring parts of the visual field, as mediated by neighboring cells in the retina. In this chapter, we will see that in primary auditory cortex (A1), the layout of cells is tonotopic: neighboring frequencies are neighbors on the cortical map. In primary somatosensory cortex (S1), we will find a map of the body’s surface—a layout called somato- topic. Taste and smell employ a different approach: their pri- mary cortical areas function by recognizing subtly different patterns rather than specific layouts.
In all the main senses, primary sensory cortex is sur- rounded by neighboring regions of cortex (secondary and tertiary) in a hierarchical relationship that leads to more ab- stract processing. Thus, the primary visual cortex (V1) is sensitive to simple lines, whereas higher visual areas respond to cars, houses, or movie stars. In the auditory system, A1 is responsive to simple tones, whereas surrounding areas encode more complex sound textures. (In Chapter 11, we will see how these areas produce language.) In somatosensa- tion, S1 responds to a stimulus on a particular spot on the skin, whereas higher cortical areas allow the deciphering of a larger sensation, such as a cat rubbing against one’s leg. In taste, the primary taste cortex communicates with the sec- ondary cortical taste area, which allows for the decoding of food type and flavor intensity. Finally, in olfaction, the pri- mary olfactory cortex receives basic information from the olfactory bulb and feeds its outputs to a network of higher areas for more subtle processing. The same general princi- ples are at work in all the systems.
Perceiving, combining, and using various information streams is the key to interacting in the world—that is, to constructing a rough picture of what’s happening outside of the body. In the coming pages we’ll learn how we detect the world “out there” (e.g., hearing, touch, pain, temperature, and chemicals) as well as the internal world (e.g., hunger, thirst, and fatigue).
Hearing Detecting sounds from the outside world is critical to sur- vival and reproductive success—whether it’s the faint crack of a tree branch that announces the presence of a predator or the cry of an offspring separated from the pack. W hen we hear sounds we are detecting vibrations carried through a conductive medium (usually air or water). The precise
06-Eagleman_Chap06.indd 165 04/11/15 4:34 pm
166 PART 2 • ChAPTER 6 Other Senses
# 158305 Cust: OUP Au: Eagleman Pg. No. 166 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
FIGURE 6.2 Pressure changes in the air are interpreted as sound by the brain. Amplitude is the size of the pressure change, and frequency (hz) refers to the number of waves that occur per second.
Outer ear
A ir
p re
ss u
re
10 Low
High
Amplitude
Wavelength
Time (s)
2
RESEARCH METHODS Psychophysics In elementary school, you likely took a hearing test in which you re- sponded to increasingly higher or lower pitch tones. What was that about? It was an example of psycho- physics, a branch of research that systematically changes physical stimuli and assesses the accompa- nying changes to your personal ex- perience, or perception, of what you just sensed (Gescheider, 1997). In other words, psychophysics (liter- ally, measuring the physics of the psyche, or mind) characterizes the relationship between external stim- uli and internal perception. By pre- senting high and low pitches and asking you to report when you de- tected something, the elementary school hearing test was using psy- chophysics to determine the physi- cal range of your senses. Similarly, when you’re judging whether a note is sharp or flat, psychophysics allows us to determine how much
the note needs to be off-key for you to detect it as such.
Over the centuries, psychophys- ics has described the relationship between stimuli and perception. As one example, imagine that you are looking at two identical spots of light. You now turn up the intensity of one of the spots until you can just notice a brightness difference be- tween the two. The minimum threshold at which this occurs is known as the difference threshold, or more commonly the just-noticeable difference. The 19th-century psy- chologist Ernst Weber was the first to notice that the size of the just- noticeable difference was system- atically related to how intense the lights were to begin with. For exam- ple, imagine the two lights began at 10 units of intensity. You may dis- cover that one of them needs to be turned up to 11 to notice a bright- ness difference. But if they both
start off at 100 units of intensity, then you would need to crank it up to 110 units. A fixed fraction of in- crease is required. This is known as Weber’s law, and the exact amount of increase required is known as the Weber fraction. Although the Weber fraction differs across the senses, what is remarkable about Weber’s law is that it holds across the sen- sory systems—whether for noticing differences in how much two items weigh, the volume of two sounds, or the size of two items.
Many useful real-world tech- niques emerge from psychophysical studies. For example, quantifying the limits of people’s senses allows for online audio and visual data to be compressed with new algorithms that throw away undetected infor- mation. This has allowed low-band- width compression schemes that are used widely in music and video streaming online.
characteristics of these vibrations—their size, shape, and frequency—determine how we perceive those sounds.
Specifically, sound travels through the air as pressure waves: increases and decreases in pressure occurring at reg- ular intervals when small segments of air are either com- pressed (increase in pressure) or decompressed (decrease in pressure) (FIGURE 6.2). Every sound is composed of several different simultaneous pressure waves, each of which oscil- lates between high and low pressure at a particular rate. The number of high/low-pressure cycles that occur per second is referred to as the frequency of the sound wave, measured in hertz (Hz). The size of the pressure change (from the peak to the trough) is called the amplitude. Frequency and ampli- tude are terms that refer to the characteristics of the pressure waves themselves. Pitch is our perception of a sound’s fre- quency (e.g., a high note or a low note), whereas loudness is our perception of its amplitude.
The sensitivity and fidelity of the auditory system are stunning. In 1935, the scientist A lvar Wilska discovered that the human eardrum could detect sounds that make it move
as little as 1029 centimeters, or the diameter of a single hy- drogen atom (Hudspeth, 1983; Wilska, 1935). Amazingly, the sensitivity of the ear is limited by the properties of air: if the auditory system were any more sensitive, it would
06-Eagleman_Chap06.indd 166 04/11/15 4:34 pm
Hearing 167
# 158305 Cust: OUP Au: Eagleman Pg. No. 167 Title: Brain and Behavior, 1e
C / M / Y / K Short / Normal
DESIGN SERVICES OF
S4CARLISLE Publishing Services
register the random movement of molecules in the air, drowning out the more information-rich signals (Sivian, 1933). In other words, hearing in humans is as sensitive as it can usefully be. Despite its exquisite sensitivity, the ear can also gauge the loudness of sounds up to 1 million times louder than the quietest sounds it can detect.
The frequency range of the ear is equally impressive— roughly 20 to 20,000 Hz in humans (Zwicker, 1961). That is, we can perceive vibrations in the air occurring between 20 and 20,000 times per second as sounds. Despite this broad range, we can distinguish between two sounds differing in frequency by only 0.8 of 1%—which translates to a note that is just barely sharp or flat (Tervaniemi, Just, Koelsch, Wid- mann, & Schröger, 2005).
In our noisy world, the air is filled with pressure waves that bombard our auditory system. The first step to actually hearing those sounds is to capture the waves themselves. We turn to that now.
The Outer and Middle Ear Capturing the sound waves is the job of the outer ear, made up of the convoluted, cartilaginous structures that feature prominently on the side of your head (FIGURE 6.3). The folds of your outer ear, the pinna, certainly look random—but they’re not. The pinna selectively amplifies certain frequen- cies of sounds coming from in front or to the side of us and selectively diminishes certain frequencies of sounds coming from behind us. The exact shape of the pinna conditions the sound coming into your ears. If you change that shape—for example, by inserting a cardboard cone in your ear—you
Outer ear Middle ear Inner ear
Pinna
Ear canalEar canal Tympanic membrane
Round window
Eustachian tube
Stapes (stirrup)
Oval window
Semicircular canals
Cochlea
Malleus (hammer) Incus
(anvil)
FIGURE 6.3 Anatomy of the ear. The pinna of the external ear captures sound and reflects it down the auditory canal. The pressure waves vibrate the tympanic membrane. These vibrations are passed by the bones of the middle ear (malleus, incus, stapes) to the cochlea, a part of the inner ear.
will have a difficult time understanding the world around you until you adapt to the new sound textures.
Because of the contours of the pinna, the orientation of the ear in relation to the sound source is crucial to how easily we can perceive a sound, its location, or even whether we hear it at all. Many animals are able to take advantage of this feature of the auditory system—consider dogs and cats, which rotate their ears toward the sound to enhance their ability to hear it. A lthough humans can’t rotate their ears in- dependently from the rest of their head, note the way that people will unconsciously cock their heads in one direction or another to capture an elusive, unidentified sound.
The deepest part of the outer ear is known as the tympanic membrane (FIGURE 6.3). This membrane vibrates like the skin of a drum—hence this is also commonly known as the ear- drum. All of the features of the outer ear are shaped to capture the pressure waves in the outside world—and the design cul- minates here, as the sound waves hit this membrane.
In the next step of the hearing process, the movement of the eardrum is mechanically transmitted to the next stage—and this job falls to the structures of the middle ear, just on the other side of the eardrum from the outer ear (FIGURE 6.3). The middle ear consists of three tiny bones, called ossicles, which are named for their suggestive shapes: malleus (hammer), incus (anvil), and stapes (stir- rup). These tiny bones are connected to one another in a mechanical chain that t
