Springer Complexity
Springer Complexity is an interdisciplinary program publishing the best research and academic-level teaching on both fundamental and applied aspects of complex systems – cutting across all traditional disciplines of the natural and life sciences, engineering, economics, medicine, neuroscience, social and computer science. Complex Systems are systems that comprise many interacting parts with the ability to generate a new quality of macroscopic collective behavior the manifestations of which are the spontaneous formation of distinctive temporal, spatial or functional structures. Models of such systems can be successfully mapped onto quite diverse “real-life” situations like the climate, the coherent emission of light from lasers, chemical reaction- diffusion systems, biological cellular networks, the dynamics of stock markets and of the internet, earthquake statistics and prediction, freeway traffic, the human brain, or the formation of opinions in social systems, to name just some of the popular applications. Although their scope and methodologies overlap somewhat, one can distinguish the following main concepts and tools: self-organization, nonlinear dynamics, synergetics, turbulence, dynamical systems, catastrophes, instabilities, stochastic processes, chaos, graphs and networks, cellular automata, adaptive systems, genetic algorithms and computational intelligence. The two major book publication platforms of the Springer Complexity program are the monograph series “Understanding Complex Systems” focusing on the various applications of complexity, and the “Springer Series in Synergetics”, which is devoted to the quantitative theoretical and methodological foundations. In addition to the books in these two core series, the program also incorporates individual titles ranging from textbooks to major reference works.
Editorial and Programme Advisory Board Henry D.I. Abarbanel, Institute for Nonlinear Science, University of California, San Diego, USA Dan Braha, New England Complex Systems Institute and University of Massachusetts Dartmouth, USA Peter´ Erdi,´ Center for Complex Systems Studies, Kalamazoo College, USA and Hungarian Academy of Sciences, Budapest, Hungary Karl J. Friston, Institute of Cognitive Neuroscience, University College London, London, UK Hermann Haken, Center of Synergetics, University of Stuttgart, Stuttgart, Germany Viktor Jirsa, Centre National de la Recherche Scientifique (CNRS), UniversitedelaM´ editerran´ ee,´ Marseille, France Janusz Kacprzyk, System Research, Polish Academy of Sciences, Warsaw, Poland Scott Kelso, Center for Complex Systems and Brain Sciences, Florida Atlantic University, Boca Raton, USA Markus Kirkilionis, Mathematics Institute and Centre for Complex Systems, University of Warwick, Coventry, UK J¨urgen Kurths, Nonlinear Dynamics Group, University of Potsdam, Potsdam, Germany Linda E. Reichl, Center for Complex Quantum Systems, University of Texas, Austin, USA Peter Schuster, Theoretical Chemistry and Structural Biology, University of Vienna, Vienna, Austria Frank Schweitzer, System Design, ETH Zurich, Zurich, Switzerland Didier Sornette, Entrepreneurial Risk, ETH Zurich, Zurich, Switzerland Understanding Complex Systems
Founding Editor: J.A. Scott Kelso
Future scientific and technological developments in many fields will necessarily depend upon coming to grips with complex systems. Such systems are complex in both their composition – typically many different kinds of components interacting simultaneously and nonlinearly with each other and their environments on multiple levels – and in the rich diversity of behavior of which they are capable. The Springer Series in Understanding Complex Systems series (UCS) promotes new strategies and paradigms for understanding and realizing applications of complex systems research in a wide variety of fields and endeavors. UCS is explicitly transdisciplinary. It has three main goals: First, to elaborate the concepts, methods and tools of complex systems at all levels of description and in all scientific fields, especially newly emerging areas within the life, social, behavioral, economic, neuro- and cognitive sciences (and derivatives thereof); second, to encourage novel applications of these ideas in various fields of engineering and computation such as robotics, nano-technology and informatics; third, to provide a single forum within which commonalities and differences in the workings of complex systems may be discerned, hence leading to deeper insight and understanding. UCS will publish monographs, lecture notes and selected edited contributions aimed at communicating new findings to a large multidisciplinary audience.
For further volumes: http://www.springer.com/series/5394 Serge Galam
Sociophysics
A Physicist’s Modeling of Psycho-political Phenomena
123 Serge Galam CREA Boulevard Victor 32 75015 Paris France
ISSN 1860-0832 e-ISSN 1860-0840 ISBN 978-1-4614-2031-6 e-ISBN 978-1-4614-2032-3 DOI 10.1007/978-1-4614-2032-3 Springer New York Dordrecht Heidelberg London
Library of Congress Control Number: 2011944508
© Springer Science+Business Media, LLC 2012 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com) To Giuliana, my so Jewish Mom (see Fig. 1)
Fig. 1 Your only freedom is... to think!
Preface
When I was at high school, I was very much interested in philosophy, politics, and history, but I was too lazy to read all the necessary books required to sustain any sound discussion. There was too much material to learn by heart along with the hundreds of great citations that each prestigious contributor had made to those fields from the early stages of human culture, which cover a few thousand years. At odds with this, mathematics was much less demanding in terms of memory and reading, and to top it all, a really exciting game. Laziness, combined with the pleasure of resolving small challenges thus put me on the scientific highway of education without a clue of where it would take me. How could I know, or even figure it out? Coming from a Jewish family expelled from Libya in the 1950s, my parents arrived in Paris without having ever met any academic in the whole of their lives, not to mention any physicist. My birth there was like a camel being put into a strange world of Martians. And sometimes, I thought that in addition, my family was from Venus. I was only instructed that to learn was the thing to do, something supposed to be much more promising than selling clothes in a little shop, at least according to Sion, my dear late father. He kept on pushing me into learning, with the irrefutable argument that I would understand later on why it was the right thing to do. Indeed, I discovered later than it is more fun to shop around looking for good bargains rather than trying to sell them. A wise man my father. And my adorable mother started calling me Professor Nimbus, due to some of my hair that pointed skywards. So, not yet being a contrarian, I kept on learning since after all it was not such a demanding thing to do, with the mathematics being quite transparent for me. It was, sometimes, even enjoyable. I thus kept on studying one year after the other until I graduated in mathematics. But proving theorems began to fade away in front of my eyes. I started to feel a bit confined by the closed world of mathematics and at the same time I actually had to wear glasses for the first time in my life. Viewing all of a sudden the so many awful details of everything around me was a shock, which in turn made me switch to physics. Do not ask me why. It is called a chaotic bifurcation. Somehow, I thought that theoretical physics would bring me closer to the real world. I agree that this was quite a peculiar view of the real world.
vii viii Preface
Doing physics, I found the idea of playing formal games to question the laws that govern basic nature rather attractive. You can have fun playing with mathematics by ignoring up to some point, part of its rigor, but always complying with the robust and nonnegotiable constraints of nature. Moreover, it was not a lonely game like mathematics. To have nature as your partner is quite a tough challenge. Maybe the fact that electronic games did not exist at that time was a key to my gaining such a vision. I thus engaged in theoretical physics, a field that is as close to nature as a Rubik’s cube! Nevertheless, rather quickly I realized that I did not have much interest in the understanding of the laws of inert matter as such. Atoms did not excite my imagination. Human behavior did. I became determined to confront the secrets of human behavior, although I had no idea of how precisely we would be able to grasp some of its big and essential features. But I was also convinced that the power and genius of the pragmatic methodology of physics could contribute towards my “crazy” plan. Of course, to implement my endeavor I had first to learn physics and in particular how physicists operate so as to identify the secrets of inert matter. My program was set. It simply took me over 40 years to highlight what stands in common between atoms and human beings. Through a long and tortuous path, “spiced” with a lot of fights as well as a huge amount of self-will, my dream finally came true. The fundamental ingredients are in this book. At least in part. And I am no longer the only contributor to the settling down of this vision into solid equations and concepts. Today, sociophysics exists as a novel emerging field of science that involves hundreds of physicists all around the world. At present, we are on the road, an unknown road with still a long way ahead of us before sociophysics is eventually validated as a science. In the future it may even become a hard science like physics, or be dismissed as an absurd pipe dream. However, for the time being, it is surely a new and exciting adventure of human research at the very frontiers of knowledge. The consequences and accomplishments are unpredictable and even unconceivable. This is all the fun and value of being engaged in fundamental research at the edge. Nevertheless, piercing some of the secrets of human behavior brings with it certain risks. That is why, while applying some of the methodology of physics to this new field of social sciences as a physicist, simultaneously, as a responsible citizen I discuss these risks thoroughly. To always keep in perspective both the dangers and the hopes of sociophysics, together with the fact that dealing with human beings is of a different nature to that of dealing with atoms, is a cornerstone requirement to allow a robust and ethical development of sociophysics. In this book, I am presenting my personal testimony, itinerary, and contributions to the field of sociophysics, but its fate will stem from the collective construction made up of the many individual contributions of researchers all over the world. In addition, the capacity to avoid a partisan politicization in formulating and addressing some fundamental issues of the political and social arena will be the key to sociophysics becoming a success story. I think.
Paris Serge Galam Contents
The Reader’s Guide to a Unique Book of Its Kind ...... xvii References...... xxiii
Part I Sociophysics: Setting the Frame
1 What is Sociophysics About? ...... 3 1.1 Science Fiction and Asimov’s Foundation Syndrome ...... 3 1.2 What Makes Sense and What Does Not with Asimov’s Fictional Psychohistory ...... 5 1.2.1 Predicting the Future Is Not Possible ...... 5 1.2.2 Historical Vs. Ahistorical Sciences ...... 7 1.2.3 Small Groups, Large Groups, Not so Simple...... 8 1.2.4 A Mathematician Could Not Have Done It, a Physicist Could, Maybe ...... 9 1.3 A Parenthesis on the Fall of Empires ...... 10 1.4 Time for New Paradigms ...... 11 1.5 Sociophysics, a Possible Novel Hard Science: Not a Zero Risk Path...... 13 1.6 Discovering the Limits of Human Freedom Opens the path to Social Freedom ...... 16 References...... 19 2 The Question: Do Humans Behave like Atoms? ...... 21 2.1 My Basic Philosophy ...... 21 2.2 Telling the Truth About What Indeed Physics Is and What It Is Not ...... 23 2.3 Physics Does Not Care About Mathematical Rigor ...... 27 2.4 Implementing a Physics-like Approach Outside Physics ...... 28 2.5 But Indeed, Do Humans Behave Like Atoms? ...... 29 2.6 Building Up an “Atom–Individual” Connection ...... 31 2.7 Our Bare Methodology ...... 34
ix x Contents
2.8 To Sum up ...... 37 References...... 39 3 Sociophysics: The Origins ...... 41 3.1 The First Days...... 42 3.1.1 Breaking the Secret of Critical Phenomena ...... 42 3.1.2 The Physicist’s Corner...... 43 3.2 The First Days, the First Fight ...... 44 3.3 From Claim to Demonstration ...... 46 3.4 The Story Behind the Scene ...... 48 3.5 More About Academic Freedom ...... 51 3.6 Surviving Within Physics by Not Playing Tennis ...... 52 3.7 Breaking the Gap with a Social Scientist ...... 55 3.8 Changing My Strategy: Back to the World of Physics ...... 57 3.9 The Secret One Shot International Seminar ...... 58 3.10 The Rising Sun of Sociophysics ...... 58 3.11 When Too Much Is Too much ...... 60 3.12 Claiming the Paternity of Sociophysics...... 61 3.13 Reorientating My Strategy Again to Join a Social Sciences Group...... 63 References...... 66 4 Sociophysics: Weaknesses, Achievements, and Challenges ...... 69 4.1 The Essential Challenges of Sociophysics...... 69 4.2 Sociophysics: A New Field Is Emerging ...... 70 4.3 Deciding the Future of Sociophysics...... 71 4.4 Sociophysics: Epistemological Foundations...... 72 4.5 Flashback to the Origins ...... 74 4.6 The Soviet-Like Rewriting of the History of Sociophysics ...... 75 4.7 Fatherhood with a Touch of Humor ...... 77 4.8 Basic Weaknesses of Growing Sociophysics ...... 79 4.9 The Positive Achievements of Sociophysics so Far ...... 81 4.10 The First Sociophysics Successful Prediction of a Precise Event ...... 83 4.10.1 Taking Risks to Validate Sociophysics...... 84 4.10.2 When the Prediction Turns Out to be True ...... 85 4.10.3 When a Prediction Fails ...... 86 4.11 Proposal to Establish a Road Map ...... 86 4.12 What the Climatologists Did with the IPCC Should Not Be Repeated ...... 87 References...... 88 Contents xi
Part II Discovering the Wonderful (and Maybe Scary) World of Sociophysics
5 Sociophysics: An Overview of Emblematic Founding Models...... 93 5.1 In a Few Words ...... 95 5.1.1 Decision Making ...... 95 5.1.2 Bottom-up Democratic Voting ...... 96 5.1.3 Terrorism ...... 96 5.1.4 Coalitions Versus Fragmentation...... 96 5.1.5 Public Opinion ...... 97 References...... 97 6 Universal Features of Group Decision Making...... 101 6.1 The Strike Phenomenon ...... 101 6.1.1 The Model...... 102 6.1.2 The Operating Mechanism ...... 102 6.1.3 The Overlap with the Physical Model ...... 104 6.1.4 The Novel Counterintuitive Social Highlights ...... 105 6.1.5 Achievements of the Model ...... 106 6.2 How Do Groups Make Their Decisions? ...... 106 6.2.1 The Symmetrical Individual Versus the Symmetrical Group ...... 107 6.2.2 The Random Symmetry Breaking Choice ...... 108 6.2.3 Anticipating the Group Choice...... 109 6.2.4 Individuals Are Often Not Symmetrical ...... 111 6.2.5 Life Is Not a Paradise ...... 113 6.2.6 Some Emblematic Illustrations of the Model...... 114 6.2.7 The Leader Effect ...... 117 6.2.8 The Overlap with the Physical Model ...... 118 6.2.9 The Model’s Achievements ...... 118 References...... 119 7 The Dictatorship Paradox of Democratic Bottom-up Voting ...... 121 7.1 The Local Majority Rule Model ...... 123 7.2 Incorporating the Inertia Effect of Being the Ruler ...... 123 7.3 From Probabilistic to Deterministic Voting...... 125 7.4 The Magic Formula for Presidency ...... 126 7.5 A Simulation to Visualize the Dictatorship Effect of Democratic Voting ...... 127 7.6 From Two to Three Competing Parties ...... 129 7.7 The Overlap with the Physical Model ...... 132 7.8 Achievements of the Model ...... 135 References...... 137 xii Contents
8 The Dynamics of Spontaneous Coalition–Fragmentation Versus Global Coalitions...... 139 8.1 The Two-Country Problem ...... 139 8.2 The Unstable Three-Country Problem ...... 142 8.3 Superposing Current Pair Bonds to Historical Ones ...... 149 8.4 From Binary to a Multiple Coalitions...... 151 8.5 The Overlap with the Physical Model ...... 152 8.6 Achievements of the Model ...... 153 References...... 153 9 Terrorism and the Percolation of Passive Supporters ...... 155 9.1 The Geometry of Terrorism ...... 156 9.2 Local Versus Global Terrorism: A Unified Frame...... 160 9.3 What Is to Be Done? ...... 163 9.4 The Various Flags of a Terrorist Group ...... 165 9.5 The Overlap with the Physical Model ...... 167 9.6 Achievements of the Model ...... 168 References...... 168 10 The Modeling of Opinion Dynamics...... 169 10.1 An Overview ...... 169 10.2 Why Is Public Opinion Often Conservative? ...... 171 10.3 The Local Majority Model and the Existence of Biases ...... 173 10.4 The Appearance of Nonthreshold Dynamics ...... 184 10.5 Mixing the Group Sizes ...... 185 10.6 Heterogeneous Agents and the Contrarian Effect ...... 188 10.7 Thresholdless Driven Coexistence ...... 189 10.8 The One-Sided Inflexible Effect and the Global Warming Issue ...... 191 10.9 The Thresholdless Case...... 193 10.10 Extending the Competition to Three Opinions ...... 195 10.11 The Reshuffling Effect and Rare Event Nucleation ...... 196 10.12 The Overlap with Physical Systems and Other Sociophysics Models ...... 197 10.13 Achievements of the Model ...... 198 10.14 In the Meantime...... 199 References...... 200 11 By Way of Caution ...... 203
Part III Democratic Voting in Bottom-Up Hierarchical Structures: From Advantages and Setbacks to Dictatorship Paradoxes
12 Highlights of the Part ...... 209 12.1 Dictatorships Can Be Democratic...... 209 Contents xiii
12.2 What It Is About ...... 210 12.3 The Wonderful World of Democracy ...... 210 12.4 Not Ruling Is Bad for You...... 211 12.5 Big Is Better ...... 212 12.6 What Matters ...... 213 12.7 The Strategic Key ...... 214 12.8 Visualizing the Democratic-Driven Dictatorship Twist ...... 215 12.9 The Key Configurations to Infiltrate a Party ...... 215 12.10 Life Is More Risky with Three Competing Parties ...... 220 12.11 Eastern European Communist Collapse Was Not Sudden ...... 220 References...... 221 13 Basic Mechanisms for the Perfect Democratic Structure ...... 223 13.1 Starting from a Naive View of Former Communist Organizations ...... 223 13.2 Setting Up the Simplest Form of a Voting Process ...... 225 13.3 The Single Random Small Group Voting Scheme ...... 227 13.4 Fluctuations, Group Sizes, and Democratic Balance...... 229 13.5 Limits of the Single Group Voting Scheme...... 231 13.6 Including Even-Sized Voting Groups ...... 232 13.7 Setting Up the Perfect Democratic Structure ...... 234 13.8 The Dynamics Driven by Repeated Democratic Voting ...... 240 13.9 Some Comments About Zero ...... 243 14 Going to Applications ...... 247 14.1 The Practical Scheme ...... 247 14.2 The Physicist’s Corner: Trying to Be a Little More Mathematical ...... 250 14.3 The Magic Formula ...... 256 14.4 What It Means in Terms of Global Size ...... 263 14.5 The Physicist’s Corner: To Make It Simpler...... 264 14.6 Putting a Limit on the Global Size ...... 268 15 Touching on a Fundamental Aspect of Nature, Both Physical and Human ...... 273 15.1 Phase Transitions and Critical Phenomena ...... 273 15.2 Revisiting the Practical Scheme ...... 276 15.3 Revisiting the Magic Formula ...... 277 15.4 Rare Dictatorial Events Versus Antidemocratic Ones...... 279 15.4.1 Another Viewpoint...... 282 15.4.2 The Physicist’s Corner...... 283 15.5 From Rare Antidemocratic Events to the Radical Efficiency of Geometric Nesting ...... 284 15.5.1 Monitoring the Rare Antidemocratic Bottom Configurations ...... 284 15.5.2 When the Radical Efficiency Turns Nasty ...... 286 xiv Contents
15.6 More About Hierarchies ...... 290 15.6.1 Randomness Is Sufficient at the Bottom ...... 290 15.6.2 Geography and Multisize Combination of Voting Groups...... 291 15.6.3 A Digression of About a Fifty Percent Score: What Is the Meaning of a Majority? ...... 294 References...... 295 16 Dictatorship Paradoxes of Democratic Voting in Hierarchical Structures ...... 297 16.1 The Inertial Effect of Being in Power ...... 297 16.2 The Dramatic Effect of Tie Break Voting in the Single...... 300 16.3 Varying the Voting Group Size ...... 302 16.4 From the Perfect Democratic Structure to the Perfect...... 305 16.5 The Dynamics Driven by Repeated Democratic Voting ...... 312 16.5.1 The Physicist’s Corner...... 313 16.6 From the Magic to the Machiavelli Formula ...... 316 16.6.1 The Physicist’s Corner...... 320 16.7 Global Size, the Practical Scheme, the Magic Formula ...... 323 16.7.1 Global Size ...... 324 16.7.2 The Practical Scheme ...... 325 16.7.3 The Physicist’s Corner...... 327 16.7.4 The Magic Formula ...... 330 16.7.5 The Physicist’s Corner...... 332 16.7.6 The Super Magic Formula...... 336 16.7.7 The Physicist’s Corner...... 337 16.8 What Happens to the Rare Antidemocratic Events? ...... 339 16.8.1 The Minimum Number of Bottom Agents to Win the Presidency...... 341 16.8.2 The Associated Number of Different Bottom Nasty Configurations ...... 346 16.8.3 The Actual Probability of a Nasty Bottom Configuration ...... 347 16.9 All Bottom Minorities and Majorities Winning the Presidency . 348 16.9.1 The Odd Case r D 3 ...... 351 16.9.2 The Even Case r D 4: The Challenging View Point ...... 356 16.9.3 The Even Case r D 4: The Running Power View Point ...... 358 16.9.4 The Physicist’s Corner...... 366 16.10 The Worrying Power of Geometric Nesting or How to Make Certain a Very Rare Event ...... 367 16.10.1 The Sudden and Unexpected Taking Over of Large Institutions ...... 368 16.10.2 The Scary Lobbying ...... 368 Contents xv
16.10.3 A Striking Idealized Illustration...... 369 16.10.4 Hint to Restore the Democratic Functioning ...... 369 16.11 Softening the Inertia Principle...... 370 16.11.1 The Physicist’s Corner...... 373 16.11.2 Three Competing Opinions: It Becomes Even More Counterintuitive ...... 374 16.12 Communist Collapse and French FN Victory ...... 374 16.12.1 Hierarchies Are Everywhere ...... 375 References...... 376
Part IV The Risky Business of Alliances in Bottom-Up Democratic Voting with Three-Choice Competition
17 Bottom-Up Democratic Voting in a Three-Choice Competition...... 379 17.1 Two Competing Parties in Short ...... 380 17.1.1 The General Frame ...... 380 17.1.2 Predicting the Results of Democratic Elections ...... 380 17.1.3 The Bottom-Up Voting Dynamics ...... 382 17.1.4 From Theoretical Principles to Reality ...... 386 17.1.5 From Reality to Implementation ...... 389 17.2 Three Competing Parties ...... 390 17.2.1 Two Competing Parties, a One-Dimensional Problem ...... 391 17.2.2 Three Competing Parties, a Two- Dimensional Problem...... 393 17.2.3 The Three-Party Bottom-Up Voting Flow ...... 395 17.2.4 The Physicist’s Corner...... 402 17.3 The Bottom-Up Voting Flow Diagram...... 404 17.3.1 The Frequent Case of Two Large Opposing Parties with a Small One in Between ...... 406 17.3.2 Some General .˛;ˇ; /Case Snapshots ...... 409 17.4 The “Golden Triangle” to Win the Presidency ...... 411 References...... 413 18 So Sorry, That’s the End of the Tour! ...... 423 19 I Thank You ...... 427 Index ...... 429
The Reader’s Guide to a Unique Book of Its Kind
While huge progress has been made in understanding and mastering inert matter as well as in manipulating biological systems, very little has been achieved with respect to human matter. The major difficulty is that we are the main part of it, making an objective investigation rather illusory. It is from this intrinsic impossibility that the use of a novel set of tools and concepts, extracted from a totally different but highly developed field of research, could turn out to be very fruitful. Accordingly, I ground the work in physics, a very successful and powerful branch of both knowledge and experimentation. Learning from a reality of a totally different nature, consisting of far less complicated entities, of “poor and limited” atoms, could be very efficient in embracing human complexity. Precisely because of this hypothesis, the work has been built without incorporating the numerous and apparently inextricable set of human attributes that we are overwhelmed with, as soon as we discuss human matter. The fact that solid regularities are observed from statistics in many human activities allows the postulation of the existence of simplicity beyond some part of complex human behavior. At least, my approach is founded on that statement. My “humans” will be very simple agents with a rather restrictive spectrum of individual features. Yet, they will be shown to exhibit complex and nontrivial behavior that provides novel insight into human complexity. The world of atoms is much more exciting and richer than previously thought up to the middle of the last century. Physics could provide an efficient framework in which to embrace a series of human puzzles. My paradoxical anchorage is that learning from atoms could provide substantial enlightenment for discovering the mechanisms behind some intriguing features of human collective organizations. Maybe human societies share common ingredients with the far less complicated entities involved in inert matter. Bridging the two extremes of known complexity, that of humans versus that of atoms, could turn out to be rather productive. Such a path provides a “trick” to be able to bypass our anthropocentric feeling of being superior and more complexly elaborated than everything else. Nevertheless, it should be emphasized that the whole approach does not pretend at a complete and precise description of human behavior. It aims at shedding
xvii xviii The Reader’s Guide to a Unique Book of Its Kind a new counterintuitive light on some basic problems of our social and political organizations. This book is therefore rather unusual in both its content and its presentation, not mentioning the style. This is why a short guide is necessary to state the general framework as well as the spirit behind the presentation. It deals with new materials, new subjects, new results, new issues, and in particular on a possible new paradigm with which to envision human behavior. The presentation is also unusual for a scientific book, in providing an initiation in carrying out research.
What It Is About
The main content is about sociophysics (Fig. 2). But, what is sociophysics? It is the use of concepts and techniques that are taken from statistical physics to investigate some social and political behavior. The topic does not aim at an exact description of the associated reality with respect to all its details. It focuses at singling out some basic mechanics which may be rather counterintuitive and in turn shedding new light on our otherwise taken for granted best sets and frameworks for political organizations. A few decades ago I envisioned the logical birth of the field of sociophysics stemming from the recognition of the high level of development of theories in condensed matter physics combined with a predictable increase in a physicist’s frustration of being confined to the study of the world of inert matter. I wrote several papers to support this statement. Most of the predictions came true and some are still waiting to do so. In particular I predicted that after the development and success
Fig. 2 Indeed, what is the book all about? The Reader’s Guide to a Unique Book of Its Kind xix of the “invasion” of other fields by physics, the phase we are entering currently, a counterreaction would form within the invaded fields to expel part of it, rather like in a decolonization process. However, I do hope that the result will be more successful than the case for countries. The paper titles are rather explicit for most of them with “Physicists as a revolutionary catalyst” [1], “Les physiciens et la frustrations des electrons”´ [2], “Sauver la nouvelle Byzance” [3], “Physicists are frustrated” [4], “Misere` des physiciens” [5], “About imperialism of physics” [6], “Should God save the queen?” [7]. I simultaneously wrote two foundational papers to establish the field as such and not just as single one shot contributions. The first one is in French “Entropie, desordre´ et liberte´ individuelle” [8]. It deals with the very ancient and puzzling paradox that arises from the principle of maximum entropy and the existence of life that is perceived as an ordered state as opposed to the ultimate fate of supposedly maximum disorder. The second one “Sociophysics: A mean behavior model for the process of strike” was written in collaboration with Yuval Gefen and Yonathan Shapir today at the Weizmann Institute and the Rochester University respectively [9]. It is a manifesto for sociophysics. In addition, it addresses the question of the occurrence of strikes in the working world, revealing some of the mechanisms behind several paradoxical facts that can be observed in strike phenomena.
Main Material from Peer-reviewed International Journals
The book is built from two different, but intricately connected parts. One is scientific and the other is philosophical. The scientific part deals with sociophysics itself, i.e., the construction of models, the discussion of their validity and their confrontation to real events. Most of the enclosed material has already been published in high ranking international research journals having international peer review procedures over a period of almost 30 years. It includes my original models as well as those made in collaboration with a few colleagues. This explains why not too many references are given to other research papers published in the last 15 years during which the field of sociophysics has gained the interest of quite a number of physicists around the world. The book also incorporates novel unpublished material, which is, however, derived directly from the published models and results. Many details and extra calculations are given which are otherwise out of the scope of a research paper.
Philosophical, Ethical, Epistemological, and Political Issues
The philosophical part of the book addresses the issues of ethics, morals, epistemol- ogy, and politics which are intimately associated to the very concept of sociophysics. xx The Reader’s Guide to a Unique Book of Its Kind
The goals and the approach of sociophysics could result in drastic changes in our lives, in particular if it turns out to be successful in its ambitious program. The question of the corresponding potential social dangers is not avoided. I am not positioning myself as an irresponsible scientist who is only excited by his ongoing research. I am fully aware that there exists a serious responsibility in engaging, implement- ing, and advocating the new field of sociophysics. I stand by it. Accordingly, these essential key issues for the future of sociophysics are addressed from the start of the book in the first chapter. Hints on how to neutralize possible misuses are given. No “stonewalling,” no “put-off.” These topics cover some very fundamental issues and although they have been much less developed and discussed in the literature, they are placed at the beginning of the book. At this stage, most of it is constituted from various comments and digressions, which express my personal opinion and do not reflect either a consensus among physicists working today in sociophysics or an elaborated framing for sociophysics, which indeed do not exist at present as a well delimited field. This is why several figures look like cartoons to emphasize that fragility.
The Whereabouts, the Injustices, and the Fate of the Creation of a New Field
I present a testimony about my more than 30 years of fighting and struggling to create the new field of sociophysics. I tell the whole story without any attempt to make politically correct statements and so emphasizing the numerous conflicts which have occurred. It is very rare to have access to the real story of the internal fight for the emergence of novel research, in particular within the so called hard sciences. I am undergoing this approach not just to advocate my particular case but for the sake of knowledge and the understanding of the universal character of the process, which, I think, is emblematic of most of those cases. Such real paths, which lead to the establishment of novel paradigms, are almost never known, as observed from the written idyllic logics found in most scientific textbooks. Once a breakthrough has been established and accepted, everyone, including those who have created it, wants to forget the initial tough rejection. Once the “disturbing and unacceptable” truth has been recognized as valid, its fringe and controversial scientist wants to become part of the established community. My story also demonstrates that scientists working at the challenging frontiers of cutting edge research are no better human beings than in any other field of social activity, having the same little arrangements to gain power, glory, and fame. The Reader’s Guide to a Unique Book of Its Kind xxi
Fig. 3 It sounds like real fun
Discovering the Fantastic and a Bit Scary World of Sociophysics
Once the questions surrounding the very nature of sociophysics have been clearly addressed, I can then define the guideline of the approach for setting the framework, the limits, and the first goals. To be able to understand, to describe and eventually to make predictions about social and political behavior and organizations is rather fascinating and exciting (Fig. 3). It also raises certain fears, with the real danger of becoming a “sorcerer’s apprentice.” I will first present some details of my personal itinerary in this field in creating the basis of sociophysics, so revealing the initial hostility of the physic’s community. Then I will comment on from my own experience of how paternity recognition is far from being natural in science as in other human activities. The real history of the first stages of sociophysics provides a very nice illustration of how sometimes the building of history can turn out to be misleading and even to falsely create politically correct paradigms at the price of modifying the truth.
An Initiation in Carrying Out Some Research “alafac` ¸on” of a Physicist
This book has two unique features, the first one being the subject and the second one being the manner in which the subject is presented. Instead of going through a classical presentation of the field of sociophysics, I implement it through an xxii The Reader’s Guide to a Unique Book of Its Kind initiation in carrying out some research as, using the same approach as a physicist. To develop the field of sociophysics, not only tools and concepts from physics are not just necessary, but even more important is to learn how to tackle a given problem, how to model it, step by step with an ongoing critical review of the results obtained. This method is designed for both nonphysicists, who want to become familiar with the sociophysic’s modelling of societies, and graduate physics students, who can learn how a physicist’s mind works. It is worth emphasizing that the content of the book does not require any preliminary knowledge of physics. Only very basic elementary mathematics is necessary to follow the developments presented in the book since all details are given, together with explanations on how and why.
The Physicist’s Corner
From time to time, depending on the particular problem being developed, I open a “physicist’s corner” to emphasize and enlighten what made the physicist’s thinking so specific, which in turn has allowed so many great discoveries. It contains illustrations of the “magic” and fantastic power of playing with mathematics, models, and experiments in order to reach a realistic goal.
Each Chapter can be Read Independently (Almost)
The content is divided into two parts. The first one contains no equations and focuses on explaining, positioning, arguing, telling stories, discussing epistemol- ogy, and revealing the real history of sociophysics. The second part thoroughly investigates one particular subject of sociophysics, the paradoxical effects of using majority rule voting in bottom-up democratic hierarchies. In particular a model for democratic dictatorship is set up to demonstrate the negative side effects of using some a priori very positive social procedures. Here, equations and explana- tions are mixed. To be more accessible, each Chapter is self-sufficient and can be read separately. Accordingly, some equations, graphs, and references are repeated from one Chapter to another, to make each Chapter independent. But a reading of all the Chapters in their natural order will be more beneficial for gaining a better comprehension of what sociophysics is, and to lead to a clearer understanding of how it works (Fig. 4). References xxiii
Fig. 4 The reader’s guide
References
1. S. Galam, “Physicists as a revolutionary catalyst”, Fundamenta Scientae 1, 351 (1980) 2. P. Pfeuty and S. Galam, “Les physiciens et la frustrations des electrons”,´ La Recherche JulyAugust, 23 (1981) 3. S. Galam, “Sauver la nouvelle Byzance”, La Recherche Lettre 127, 1320 (1981) 4. S. Galam and P. Pfeuty, “Physicists are frustrated”, Physics Today Letter April, 89 1982) 5. S. Galam, Misere` des physiciens, Pandore 18, 57 (1982) 6. S. Galam, “About imperialism of physics”, Fundamenta Scientiae 3, 125 (1982) 7. S. Galam and P. Pfeuty, “Should God save the queen?”, Physics Today Letter October, 110 (1983) 8. S. Galam, Entropie, desordre´ et liberte´ individuelle, Fundamenta Scientiae 3, 209 (1982) 9. S. Galam, Y. Gefen and Y. Shapir, Sociophysics: A mean behavior model for the process of strike, Journal of Mathematical Sociology 9, 1 (1982)
Part I Sociophysics: Setting the Frame Chapter 1 What is Sociophysics About?
Sociophysics deals with one of the most ancient of human dreams, which is simultaneously its nightmare, i.e., the capability to predict and thus to control human behavior. As is often the case, science fiction creates what science will be able to achieve in the future, preceding in part the reality to come. But while succeeding in embodying some of the features of the future, it usually fails to grasp its more fundamental aspects, which is somehow comforting as regards our nonprewritten destiny, which always has many uncertainties.
1.1 Science Fiction and Asimov’s Foundation Syndrome
A few years ago, after giving a lecture about some of my work on sociophysics, once the audience’s questions were over, on my way out of the auditorium, someone came to me with big shining eyes to ask me if I knew about Hari Seldon, his psychohistory, Foundation, and Asimov’s trilogy? This scenario repeated itself a few times in a row. Each time I was surprised and my answer was polite but somehow negative. I had never heard about this Asimov and his program before, and I think I was a bit reluctant to consider that I might be practicing some science fiction-like ideas. But all these people were telling me in very convinced tones and with great enthusiasm that I should read the novels, since somehow I was doing what the novel’s main figure, Hari Seldon, does in the Asimov saga. According to my questioners, my sociophysics was identical to his psychohistory. One more step and I would have become the incarnation of this novel’s figure; this was a strange feeling for a physicist whose aim was and still is to create a solid science of social behavior. To appear as a novel’s character was thus a disturbing threat that risked jeopardizing the seriousness of my approach. Just after these repeated events I went to America to an International Conference and I decided that it was time to check out what was behind these strange and
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 3 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 1, © Springer Science+Business Media, LLC 2012 4 1 What is Sociophysics About?
Fig. 1.1 Hari Seldon is watching me
intriguing statements about both my work and my identity. I went to a large bookstore, wearing large wrap-around sunglasses to buy incognito Asimov’s first Foundation trilogy (Fig. 1.1). After an exciting read, I ended up being a bit disappointed, but secure as regards my own construction. First, I understood why these people were making the analogy with me. Then I identified what was similar, indeed very little, and what was different, the main part. However this comparison turned out to be very fruitful since it allowed me to identify the possibilities and the impossibilities associated with the IDEA of quantifying human behavior. For the sake of both curiosity and clarity, let me first review in short what The Foundation is about. Isaac Asimov is a well known and successful science fiction author who has created many interesting and intriguing worlds that could possibly come true in the future. Among the numerous books he has written stands the trilogy called “Foundation” with Foundation [1], Foundation and Empire [2], and Second Foundation [3]. The central figure is a fictional character called Hari Seldon, a mathematician living in some Galactic Empire. In the course of his research, he envisioned a new field from mathematics that he called “psychohistory.” The radical content of this novel theory was to embody the possibility of predicting the Galactic Empire’s future. Thanks to the size of its full population, consisting of billions of individuals, psychohistory was supposed to be able to make predictions using an analogy with the statistical theory of gases. At first, Seldon had only the formal proof that his program was feasible in principle, but he had no clue as to how to implement the theory in practice. He emphasized this limitation while publicizing his theory at a mathematical conference. Indeed he was pessimistic at that stage about the fact that the practical 1.2 What Makes Sense and What Does Not with Asimov’s Fictional Psychohistory 5 aspect could not be resolved in the near future. The technical difficulties were tremendous, even if the general theory was mastered. His incredible program was not yet near to becoming a real scientific field. However, already some politicians became interested and / or scared by the power of the theory. They were proven right since later in the novel, Seldon succeeded in overcoming the technical barrier that made it possible to turn the theory into an efficient predictive political tool. With the capacity to develop psychohistory into a practical set of equations, he was able to predict the Empire’s future in terms of probabilities. His prediction was the eventual fall of the Galactic Empire. The whole saga follows from that statement which appeared to be accomplished later in the novel.
1.2 What Makes Sense and What Does Not with Asimov’s Fictional Psychohistory
Of course Asimov’s psychohistory is a fiction created by a science fiction author and therefore it could seem out of place to discuss it in a “serious” scientific book. However, it is not superfluous to comment on it since sometimes science fiction, like art, is capable of picking out some features of what will become reality later on. Obviously, not every dream becomes reality, but some do, at least in part. Simultaneously, some realities produce dreams. Let us thus examine what makes sense and what does not in Asimov’s vision of the role of mathematics in the course of collective and individual human lives.
1.2.1 Predicting the Future Is Not Possible
Using an analogy with the statistical theory of gases, psychohistory is supposed to be able to predict the future of the whole population of the Galactic Empire consisting of billions of individuals. It relates to the fact that dealing with large numbers of particles results in the averaging out of the initial conditions to result in the existence of macroscopic laws which govern the collective behavior of the ensemble of particles. This is the way it works for gases. The question is then (Fig. 1.2): Given the hypothesis that it was possible to collect all the information concerning the individuals of a large population, would it become possible to describe its future using some macroscopic quantities and their associated laws of evolution? In other words, imagining it being possible to store in a huge computer at a given time all the individual characteristics of every person from the whole of the world’s population, knowing in addition their laws of interaction, does this make it possible to calculate, and thus predict, their evolution with time, i.e., their future, their history to come? 6 1 What is Sociophysics About?
Fig. 1.2 Predicting the future as against predicting an event: it is not possible to predict the future but sometimes it is feasible to predict an event
Although the question sounds right, the answer is “no,” it will never be possible to calculate their future however powerful the computer and however large is the quantity of personal information collected for each person. What is erroneous is the mere idea of predicting the future of a population, the future being taken in terms of its coming history which implies a large spectrum of levels and contents (Fig. 1.3).
Fig. 1.3 Science fiction versus science: it sounds right, it is almost right, but it turns out to be wrong
The analogy with a gas does not hold since in physics the control of the evolution of a gas presupposes that the gas is contained in an enclosed space and constrained by known external parameters. In contrast, an assembly of people can perform a large number of activities that are carried out under unknown external conditions. 1.2 What Makes Sense and What Does Not with Asimov’s Fictional Psychohistory 7
However, the analogy can be sustained if the range of possible activities by a group of people is reduced to just a few issues and occurring over a finite interval of time, focusing on one particular event. Then prediction becomes possible at least in principle and under certain conditions. It is precisely in this context that sociophysics could be developed. It is far less ambitious but yet an incredible challenge. A society is not a perfect gas. Even most of inert matter is not a perfect gas. Physics does not predict the evolution of all inert matter. It separates things, isolates levels, controls the external conditions, and defines orders of magnitude in time and energies to discover the appropriate laws relevant to each subcase of the incommensurate number of available atoms.
1.2.2 Historical Vs. Ahistorical Sciences
To elaborate on the respectively different natures of historical and ahistorical sciences is out our scope here. Nevertheless let us mention the essential separation between them. The first ones are concerned with unique objects, such as the universe, the planet earth, its climate, myself, you, and so on. The second group allows a hard science approach since it concerns a collection of identical objects whose past does not matter to its evolution with time. The history becomes irrelevant and thus the future cannot be defined as such. The social and political phenomenon driven by the so-called global warming is a perfect illustration of these difficulties, impossibilities, and wrong statements since although climatology is not a hard science, it is instrumented by climatologists to make unfounded and usually catastrophic predictions, claimed to be scientifically proven. I wrote a book on this “hot” subject. Unfortunately, at present (Spring 2011), it is available only in French [4]. And even more unfortunate, I have the feeling it will stay so. I also published a research paper in the international journal of physics, Physica A, about the dynamics of public opinion related to several issues driven by incomplete scientific data such as that involved in global warming, evolution theory and H1N1 pandemic influenza [5]. To conclude, I do not think history could be predicted even in principle, given our current tools of research and perception of the world. But it is worth stressing that predicting the collapse of an empire is not predicting the future. However localized in time, events may well be predicted. It is unfortunate but necessary to clearly state that what could count as a “global science,” which embraces multiple aspects of different sciences at various scales to study a given phenomenon such as the climate, still does not exist. For instance, physics is not a global science, with its well separated fields of knowledge that include high energy physics, nanophysics, condensed matter, astrophysics, and others. 8 1 What is Sociophysics About?
1.2.3 Small Groups, Large Groups, Not so Simple
Statistical physics deals with averages that make use of thermodynamic limits. Because the number of objects tends toward infinity, an essential ingredient is to ensure a correct mapping between the various theories and the corresponding experiments. Such a procedure is necessary to average out microscopic fluctuations. But in some situations fluctuations turn out to be essential, like in the so-called phase transition in which a system changes its macroscopic order from one organization into another. For our purpose, what matters is to underline the fact that increasing the number of particles reduces the importance of the initial conditions, making the evolution independent of them. For a large enough system, whatever the initial conditions the equilibrium state will be unique. This situation does not apply to the case for a small number of particles. Here stands the mathematical basis behind Seldon’s requirement for a huge population in order to get his psychohistory to work, which is indeed the case with the Galactic Empire. However, the larger a system is, the more probable is the occurrence of an extremely rare local event which can totally destroy the otherwise predictable and normally accomplished evolution. Asimov actually introduced this ingredient with the appearance of a “mule” in his saga. And as expected, this “mule” seriously shook all of Seldon’s previsions. Therefore, this feature of the Asimov panoply makes sense. I encountered this phenomenon while developing some modeling of opinion dynamics. Extending my work on democratic dictatorships [6, 7] to opinion dynamics with a group of Swiss colleagues, Bastien Chopard, Alexandre Masselot, and Michel Droz, we found that the opinion formation obeys threshold dynamics [8]. For two competing opinions A and B, an unstable threshold determines a full range of values for initial supports for which an A victory is guaranteed through a public debate and others for the B opinion victory. Later on, performing large simulations with Bastien Chopard and Michel Droz, we were surprised to discover the existence of what we called “killer geometries” [9, 10]. Given a reaction/diffusion like model it appears that some peculiar local geometries could emerge with the astonishing property of having a nonzero probability of invading the whole population. A few elements adequately organized can thus win against a whole population. They do not even constitute a small minority, being simply a few individuals. With Jan Radomski, a Polish colleague, we then extended this above finding to propose a mechanism for cancerous tumor growth. In particular, we emphasized that while such an occurrence is extremely rare, it becomes very probable after a long period of life. Our paper is entitled “Cancerous tumor: The high frequency of a rare event” [11]. From our model, natural survival from cancerous tumors, i.e., using the body’s natural defenses, becomes random after a certain age. At the opposite extreme, at younger ages, survival to cancer cells is almost certain. However, these conclusions were not extracted from real data but only from a model, which is far from being close to the real evolution of cancerous tumors over time. 1.2 What Makes Sense and What Does Not with Asimov’s Fictional Psychohistory 9
It is therefore of central importance to underline that deterministic laws can perfectly exist with the possibility of “mutants” which eventually break up the prediction. The above example of a dynamics threshold with the existence of killer geometries is a good illustration, which could also provide a hint in helping to understand lobbying effects.
1.2.4 A Mathematician Could Not Have Done It, a Physicist Could, Maybe
If the “mule” event made sense, the character of the main figure of the novel does not. Choosing a mathematician to create psychohistory is not a valid setting for the fiction, in the case where one tries to take it as a solid scenario. To make my point, and to be a bit more provocative, I would say that a mathematician could not have invented physics. By a mathematician, I mean a mathematician doing mathematics. Of course a mathematician becoming engaged in another field is a different matter, but then he or she would no longer be a mathematician. To support this surprising argument, I could mention the mathematicians working in economics, who keep playing with mathematics and not much with economic reality. If necessary, the 2008 financial and economic crisis could be used to illustrate my point. Accordingly, if a mathematician could not have created physics, and if mathe- maticians have created the field of economics, which is so far from the real economy, how could a mathematician come up with a psychohistory-like theory which will have a connection to real history? Moreover, designating history as a social field that can be quantified is also absurd. By its very nature, history does not allow a predictive methodology since it is the unique history of a unique system. I do not want to be misunderstood. I am not taking a contemptuous posture toward mathematicians. Without them, there would be no physics. I simply want to stress the radical gap that exists between the respective frameworks of the minds of mathematicians and those of physicists. Physicists use mathematics as a very useful and powerful tool to investigate the hidden laws of inert matter. What matters is reaching a conclusion that corroborates with experimental results. If, while building up a theoretical model to explain some phenomenon, mathematical rigor gets in the way and prevents the reaching of the required result, the physicist will have no “moral” or professional qualms in making up a mathematical hypothesis as long as it allows the bypassing of the difficulty. On the contrary, often a physical explanation will be created within which the mathematical distortion is given an explanation that has yet to be validated. To a physicist’s mind, what prevails is the result, not the mathematical rigor of the demonstration (Fig. 1.4). The experimental level always prevails over the mathematical level. I could formulate the difference in another way by stating that given a problem, for a physicist, what matters is to find some minimum conditions in order to obtain a solution. On the opposite side, for the mathematician, what matters is to prove the universality of the solution. 10 1 What is Sociophysics About?
Fig. 1.4 Who could eventually predict social and political events? Mathematicians cannot. Physicists might be able to
This is why only physicists could apply mathematics to describe another field of the real world in a quantitative way, analogous to what they do in physics. Neither chemists nor biologists could do it because both of their fields deal with the modification of the nature of their objects. In the study of human behavior, humans are assumed to remain unchanged, as with atoms. When mentioning mathematicians and physicists, I am not concerned about the education or knowledge people may have but about their cognitive framework for tackling a problem. In principle, a mathematician has a very different cognitive nature than a physicist and vice versa.
1.3 A Parenthesis on the Fall of Empires
The core of Asimov’s trilogy is the prediction by the mathematician Seldon, ahead of time, of the fall of the Galactic Empire using his psychohistory, i.e., a set of equations. Taking his prediction as a solid fact and considering that it could not be avoided, he engages in a race to save all of the existent human accomplishments in order to shrink drastically the time that would normally be required to rebuild a new and better universe. Instead of the new humanity starting from scratch, Seldon envisions having delivered to it all of the accomplishments of the previous humanity. What a noble cause, isn’t it? But for me, it is the prediction content which is the most interesting. It is amazing to see that the prediction of the fall of an empire is considered as a dangerous scoop, a subversive fact that must be kept top secret. It is indeed incredible to notice that while through known past history every and each empire has fallen, every and each new empire is convinced that it is eternal, with the self-certainty of constituting the ultimate state of history and the fate of existing forever. People tend to forget that history demonstrates that every empire’s fate, whatever the height of prosperity it achieves, is to fall, collapse and then to disappear rather quickly (Fig. 1.5). 1.4 Time for New Paradigms 11
Fig. 1.5 The destiny of any empire is to eventually fall
Any empire is going to fall at some time; that is the remarkable evidence, which is, however, always forgotten while living in any particular empire. Such a situation could result from the fact that to sustain respect toward an institution’s authority, the belief in it lasting forever is a prerequisite. Although one can be certain that one day the institution will fall, the calculation of the particular date of its fall is always unknown. Knowing the date in advance would clearly offer plenty of strategic advantages to the person possessing the information. Such a hypothetic possibility would provide tremendous consequences and advantages to the ones who knew.
1.4 Time for New Paradigms
The idea in pointing at connections between on the one hand, collective phenomena in physics, and on the other, collective behavior in human systems, appears today as being rather appealing, almost as a natural link. That was not the case 30 years ago when paradigms for physics and human societies were so far apart. Nevertheless, in the late 1970s, physics was the very unique field that was able to tackle theoretical and experimental challenges arising from the existence of collective phenomena. A great deal of new materials yielding exotic features have been discovered and understood, thanks to the development of the appropriate concepts and techniques. The physics of collective phenomena is now a well 12 1 What is Sociophysics About?
Fig. 1.6 Our present “glasses” are insufficient to see the world
established and flourishing field of research, particularly in its massive and smart use of the enormous developments in simulation capabilities triggered by the ongoing increase in the power of computers. The field of collective phenomena is a major ingredient for understanding inert matter, and in turn for discovering novel and unexpected properties such as the so-called high Tc superconductors, which opened the way to dream about the possibility of building superconductors at temperatures not too far from room temperature in the near future. The massive recent development of nanotechnologies is another example. Nowadays, globalization has made economic, financial, social, political, and religious collective phenomena the focus of the most striking world events. This state of affairs demonstrates the richness as well as the many different fears associated with essential contributions to all new developments at every level of human society. And yet many innovations are still to come. The supremacy of collective phenomena nowadays is often perceived as being negative, preventing the search for one’s own personal accomplishment. Individualism does not seem to get along with the reality of world collectiveness. You find the same shops, the same garments, and the same food all over the world. At the end of the last century, we shifted from a world of stability to a world of chaotic and probabilistic dynamics with incessant changes in the collective and individual reality of everyone’s lives. These features are usually perceived as catastrophes. Many would like to go back to the “old” world to suppress all the surrounding and internal disorder and to recover the happiness of an ordered world and life. But to go back to a past which no longer exists would be very counterproductive.Indeed, we should highlight the richness and advantages of chaos and disorder in order to exploit it in every part of our lives. However, to implement such a counterstatement, we desperately need to create new tools and new paradigms so as to apprehend collective phenomena as part of our basic resources for the future (Figs. 1.6 and 1.7). All our concepts and ideals are outdated, with the most recent ones coming from the nineteenth century while our world has drastically changed. The physics 1.5 Sociophysics, a Possible Novel Hard Science: Not a Zero Risk Path 13
Fig. 1.7 We need appropriate “glasses” to see the full world
of collective phenomena is a promising reservoir for innovative concepts and representations for addressing the tremendous new challenges facing our worldwide society. This book lays the groundwork for such a program, setting the foundations for sociophysics. It is a contribution to why and how the physics of collective phenomena can provide new keys to the global political and social world. Therefore, we address the fundamental question of “do humans behave like atoms?” to which the answer is “yes in certain parts of social and political collective behavior” and “no in other aspects of individual behavior” (Fig. 1.8).
Fig. 1.8 The book’s basic DoDo humanshumans bbehaveehave llikeike aatomstoms question
ThatThat isis thethe BIGBIG questionquestion
1.5 Sociophysics, a Possible Novel Hard Science: Not a Zero Risk Path
However, after making such a self-confident statement as stated above about the feasibility of quantifying some of the human collective behavior, one should immediately restrain the excitement with a fundamental basic note of caution. 14 1 What is Sociophysics About?
Fig. 1.9 Not a zero risk adventure
While the need for new concepts is clear, and the temptation to pick up new paradigms from the physics of collective phenomena is rather attractive, it may turn out to be both misleading and dangerous. Such an approach should indeed be carefully controlled (Fig. 1.9). To just map a physical theory built for a physical reality, onto a social reality, could at the best be a nice metaphor, but without the possibility of predicting anything, and at the worst, lead to a misleading and wrong social theory. It may as well become a false argument of authority in order to impose political views. Since it is immediate and easy to develop a series of a priori metaphoric mappings between on the one hand, magic words and pictures from physics, and on the other, problems and issues in the social world, it may be very tempting to create an artificial appearance of scientific arguments to impose subjective and political views in the name of science. This idea could lead to dangerous social manipulations. I could mention here again as an illustration the global warming phenomena about which climatologists claim to have scientific proof but which has indeed become a political view of the world. It is a mere vision, that is politically correct and fits to the desire for order to recover the illusion of understanding and controlling of the world’s development through the illusion of mastering the evolution of planet Earth (Fig. 1.10). Our task is to borrow from physics those techniques and concepts that can be used to build a collective theory of social behavior, but within the specific constraints of the psychosocial reality. Another danger for the physicist is to stay in physics, using a social terminology and a physical formalism. On this basis, one should be aware of such dangers and account for them while becoming engaged in this new and exciting approach to the description of social phenomena. 1.5 Sociophysics, a Possible Novel Hard Science: Not a Zero Risk Path 15
Fig. 1.10 A rather promising path
The contribution from physics should thus be restricted to guidelines for the modeling of social realities. Such a limitation does not make the program less ambitious. On the contrary, it opens very solid perspectives for efficiently tackling the current ongoing process of globalization. In fact, it may even oppose social manipulations. In addition, given the level of current and future world problems, our goal is twofold; firstly to account for the inherent risk to sociophysics of keeping a constant effort to maintain it at a minimum, and secondly, to proceed in the elaboration of a new pair of “glasses” to improve our view at the world scale. On top of the above risks, and once it has been decided to take responsibility for them, another difficulty arises; that of the ethical issues. These issues have to be addressed before taking the proposed path. It is rooted in the very nature of the sociophysics approach and in particular, to its fate. After some amount of research and work, sociophysics will eventually be proved to be either wrong or right. In the first case, it would have been worth trying it, and then the whole thing would have to be forgotten as being just another unfounded human dream or nightmare. That is all. Moreover, all the people who were scared of the adventure from the beginning, would be reassured and reinforced in their conviction that mankind does not obey an equation. It would be an indirect proof that human beings really do enjoy individual free will and are free to decide their collective fate. In short, human beings and societies have nothing in common with atoms and inert matter. Human superiority would thus be reinforced by the failure of the attempt. On the contrary, in the case of the second possibility, of sociophysics being proven to be well founded, the core hypothesis stating that human beings at a social level do obey “natural laws” in the same way that atoms do, would be validated. Therefore, besides the shaking of the human dream of superiority over other species, it could, at least in principle, pave the way to possible political manipulation. Instead of using wrong arguments, lies or coercion, it could be enough to apply a set of equations to control a population in a subtle and invisible manner. It would validate the myth of “Big Brother” with all its underlying fears (Fig. 1.11). That is the pessimistic conclusion. A more cheering alternative also exists. 16 1 What is Sociophysics About?
Fig. 1.11 Physics may provide new tools to help us catch up with the twenty-first century
1.6 Discovering the Limits of Human Freedom Opens the path to Social Freedom
Within the above contradictory possibilities, on the one hand is the optimistic hypothesis of the possibility of a hard social science, and on the other, the pessimistic idea of the lack of social freedom. The question that immediately arises is, who would be the “Big Brother” and for what purpose would it be used for? The question is formulated in a unique way but the answer is plural. If the theory is proven to be true, it could be applied for the benefit of various social and political groups. It sounds indeed very frightening, but fortunately while the objection is solid and legitimate, it is effectively socially irrelevant. On the contrary, proving that social freedom is an erroneous myth would open the path to achieving an effective freedom (Fig. 1.12). Today many very sophisticated and efficient tools are available and used to influence people, such as in advertising, to customize consumers and voters. There exists a well developed empirical knowhow for activating social and economic manipulation, that will become more and more refined with ongoing improvements in performance. But this apparatus of knowledge and tools is not known to most of the people who are being influenced. Manipulating people while they are convinced of their freedom and freewill renders hopeless any attempt to oppose it. Control and conditioning are already very active, although within both an informal scheme and a wrong social representation of social freedom. 1.6 Discovering the Limits of Human Freedom Opens the path to Social Freedom 17
Fig. 1.12 Today, manipulation is very efficient but everyone believes in spontaneous freewill
Singling out objective laws of human behavior will clearly set borders to the existing, although invisible, limits of human freedom. Simultaneously, it will open up our minds to the ways in which we are being manipulated by both external forces and current wrong conceptions. Understanding our self-limits and their mechanisms will produce a solid framework for overcoming the limitations, and thus create the possibility of more freedom. To illustrate my argument, let us look at another similar problem which occurs in the physical world: that of gravity and of the “freedom” for man to fly. For many centuries, men dreamt of being able to fly like birds do. Mixing all kinds of convictions, esoteric beliefs, and ad hoc devices, many attempts were undertaken to do so, with always the same dramatic failure of crashing to the ground. At some point, that dream of flying had to be given up with the discovery of the law of gravity, which yielded the scientific proof of the impossibility for men to fly. Due to the attraction from the earth, men were forced to stay stuck on the ground (Fig. 1.13). However, this restricting proof opened the way to overcome this fundamental limitation to our physical freedom of movement. Knowing exactly why man cannot get off the ground with the singling out of our limits, has allowed the design and the production of airplanes with the capability of flying. It has been the understanding of the natural limits set to our freedom of movement, which in turn has driven an increase of that freedom pushing away the corresponding limits. On the contrary, having the illusion of a freedom, which does not exist, is misleading and sometimes could turn dangerous. Along this line, it is the finding of eventual limitations to our free will that will eventually allow us to overcome them. It is my position here to assume that the singling out of the eventual social laws which constrain our free will is the prerequisite to the setting of more social freedom. The subtle assumption behind my 18 1 What is Sociophysics About?
Fig. 1.13 Science could single out what restrains our freedom, besides the physical limitations philosophy is that it is paradoxically the very existence of individual freedom which in turn produces the collective lack of freedom by implementing wrong conceptions for the building of social organizations (Fig. 1.14).
?
! ?
!
Surface of the earth
Fig. 1.14 Once the limitations to our freedom have been discovered, science could at least, in principle, shake up the framework that reduces these very limits References 19
References
1. I. Asimov, “Foundation”, ISBN 0-553-29335-4, Gnome Press (1951) 2. I. Asimov, “Foundation and Empire”, ISBN 0-553-29337-0, Gnome Press (1952) 3. I. Asimov, “Second Foundation”, ISBN 0-553-29336-2, Gnome Press (1953) 4. S. Galam, “Les scientifiques ont perdu le Nord, Refexions« sur le rechauffement« climatique”, Plon, Paris (2008) 5. S. Galam, “Public debates driven by incomplete scientific data: The cases of evolution theory, global warming and H1N1 pandemic influenza”, Physica A 389, 3619 (2010) 6. S. Galam, “Majority rule, hierarchical structures and democratic totalitarism: a statistical approach”, Journal of of Mathematical Psychology 30, 426 (1986) 7. S. Galam, “Social paradoxes of majority rule voting and renormalization group”, Journal of Statistical Physics 61, 943 (1990) 8. S. Galam, B. Chopard, A. Masselot and M. Droz, “Competing Species Dynamics”, European Physical Journal B 4, 529 (1998) 9. B. Chopard, M. Droz and S. Galam, “An Evolution Theory in Finite Size Systems”, European Physical Journal B 16, Rapid Note 575 (2000) 10. S. Galam, B. Chopard and M. Droz, “Killer geometries in competing species dynamics”, Physica A 314, 256 (2002) 11. S. Galam and J. P. Radomski, “Cancerous tumor: the high frequency of a rare event”, Physical Review E 63, 51907 (2001) Chapter 2 The Question: Do Humans Behave like Atoms?
The analogy, if any, between men and atoms is discussed to single out what can be the contribution from physics to the understanding of human behavior. The basic assumptions of the approach of sociophysics are formulated together with the methodology used to tackle a given problem taken from the social and political worlds [1, 2].
2.1 My Basic Philosophy
Once ethical questions and feasibility have been clarified, we can proceed in making our scheme for building a new approach to human behavior more precise, including the psychological, social, economic, and political aspects. It should be strongly stressed that the goal is not to substitute a physical view to all aspects of human life. As a first basic step, we aim to bring to light the very plausible existence of quantitative laws which govern human behavior. In a second step yet to come, sociophysics could turn into a quantitative science in order to discover the actual laws of social behavior in the same way that physics did so with the laws governing inert matter. It does not mean that the respective laws are identical. We are convinced that understanding the laws of human behavior can only be a benefit to human life and to humanity. But as mentioned above, in order to avoid a dangerous misuse, it is necessary to emphasize my methodology before elaborating on my different studies of social phenomena. The same procedure will always be used whatever the problem investigated. Thus, hypotheses and associated weaknesses will be systematically and clearly stated to set the limits of the model and the corresponding relevance to reality. Sociophysics does not claim to reach an exact description of a human group, but instead aims to shed new light on human phenomena, which are otherwise so complicated that any assumed truth is by nature misleading, and eventually wrong.
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 21 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 2, © Springer Science+Business Media, LLC 2012 22 2 The Question: Do Humans Behave like Atoms?
Fig. 2.1 Current paradigms of society belong to the physics of the ninetieth century
Behind the use of the word “physics” in our so-called “sociophysics” stands the way physics proceeds in seeking to understand the laws of inert matter, rather than the use of the laws of physics themselves. This particular way relies on the modeling “a` la fac¸on” of the physicist, the physicist’s way of modeling (Fig. 2.1). At the core of our underlining philosophy is the view that facing such a complex and complicated world such as that of human societies, only an over simplified vision of it can indeed capture a substantial part of its essential features. The initial step is to constitute the first essential presuppositions and basic foundational assumptions of a science that has yet to be set out. This statement could sound like an epistemological contradiction, but only in appearance, as will be clear later in becoming acquainted with the physicist’s frame of mind (Fig. 2.2). To proceed with this working hypothesis, the first main path to take is to create “artificial” worlds in which everything is controlled, i.e., in which all the ingredients are clearly defined, the laws of interactions are explicitly written and the range of variations of all the parameters are set, together with the definition of the eventual dynamics of evolution. Once all that is done, the associated collective properties can be studied to find out how “artificial” humans behave within the given “artificial” world. Once this procedure has been carried out, we will proceed in checking out if some of the discovered features, which are active within our “artificial world,” could mimic some of the features of the real world we aim to capture. In case it does work, we will not jump to the conclusion that we have discovered some “social truth” but we will suggest that the mechanisms used to describe the “artificial world” may be very similar to the ones behind the corresponding real social feature. We believe that it is from simplicity that theoretical complexities can appear in a way that allows subsequent understanding of the real and many complex 2.2 Telling the Truth About What Indeed Physics Is and What It Is Not 23
Fig. 2.2 Physics may provide useful novel paradigms to help us tackle the challenges of the twenty-first century aspects of the social world. The permanent challenge is to build a model to perform calculations in order to obtain data and results, which in the second step can be compared to the real counterpart of the phenomenon, either via the setting up of experiments, as in physics, or via quantitatively observed trends from the social world. On this basis, the model is either validated or invalidated, with the awareness that we are building up a model by successive layers, each layer being an additional level of complexity to mimic the real system.
2.2 Telling the Truth About What Indeed Physics IsandWhatItIsNot
It is of importance to emphasize that contrary to the general belief, physics from the start, gave up the search for an absolute global and unique truth to describe inert matter. By “the start,” I am referring to the early origins of the modern scientific method set in the fifteenth and sixteenth centuries which are now used in physics. Man as a subjective body driven by beliefs was separated from the performing of an experiment, which was required to yield the same results independently of where and by whom it is conducted. The establishment of an independent theoretical corpus was a key to the development of science in parallel to the experimental checks of its predictions. The genius of physics has been its success in separating various levels of the inert world in terms of energy and time scales such that it became possible to investigate the inside of each one, gaining in turn substantial insight to make solid 24 2 The Question: Do Humans Behave like Atoms?
Fig. 2.3 The iterative path of physics-like modeling predictions. Physics does not provide one single view of all aspects of inert matter. It discriminates between high energy physics, nuclear physics, atomic physics, condensed matter physics, astrophysics, cosmology, and more. While mastering aspects of each part of the underlying truth, the possibility of the full truth is de facto abandoned although some are still dreamt of, such as in achieving the so-called “grand unification” of the four forces. This well proven and powerful operative mode articulates via the design of so called “toy models,” which are by nature extremely crude with respect to the reality they address. But indeed, although far removed from it, they have been capable of exhibiting nontrivial results which have been shown to be shared by the corresponding real world. Once, a “toy model” is defined, solved and validated, it becomes possible to enrich it with additional ingredients in order to narrow the gap with reality and cope with it in a more realistic way. It is this very “ try and check” procedure which has led to the many breakthroughs accomplished in science (Fig. 2.3). We can illustrate the method by considering some system built from hypothetical agents denoted XY3Wz, which could be identified with their different attributes as shown in Fig. 2.4. These agents evolve in assemblies. If we wished to study their 2.2 Telling the Truth About What Indeed Physics Is and What It Is Not 25
Fig. 2.4 A system constituted with hypothetical agents denoted XY3Wz behavior, since they are too complicated to be solved at once, a physicist will start by greatly simplifying the object by selecting only one of the agent’s attributes. Such a procedure could lead to several different options, each one focusing on one particular attribute. Three different possibilities are shown in Fig. 2.5.Moreare possible. The simplified agent can then be placed in an assembly whose properties are investigated. This shows why different models are often invented to describe the same problem.
Fig. 2.5 The first step to model the XY3Wz problem is to simplify the attributes of the agent so as to obtain a simplified version of the XY3Wz agent. Three different versions are given. At this stage, they are all on the same footing, i.e., very far from the real agent
In the second step, the simplified XY3Wz agent is used to study an assembly, as in Fig. 2.6. The results obtained are then compared to the available data from the real system. Most of the time, a substantial gap is found between each model and the real system. 26 2 The Question: Do Humans Behave like Atoms?
Fig. 2.6 Three different models of an assembly of the same XY3Wz system
On this basis an improved version of the agent is required. Usually, this step forward is achieved by adding one attribute to the simplified agent in order to obtain a more elaborate model. Figure 2.6 shows an example of adding one attribute to each one of the three simplified agents. The above process can be repeated, combining results from the different models so as to eventually reach a single model which is closer to reality (Fig. 2.7).
Fig. 2.7 One attribute is added to each one of the three simplified XY3Wz. Then crossing the various models, a solid one can eventually be obtained which includes all the attributes of the XY3Wz agent 2.3 Physics Does Not Care About Mathematical Rigor 27
2.3 Physics Does Not Care About Mathematical Rigor
At this stage, we need to emphasize one additional fundamental feature of the nature of physics. While the use of modeling in physics has been tremendously powerful in establishing the field as an exact hard science, capable of building concrete and efficient experimental devices, its power comes from the empirical use of mathematics to describe real phenomena. This means that it is not the mathematical rigor that prevails but the capability to reproduce particular properties using some mathematics. It is the exact opposite of what economists have been doing for decades, who have focused on the mathematical rigor of their model rather than their ability to reproduce real features. Another essential characteristic of physics is that all the results obtained from the various models are aimed, sooner or later, at being tested against experimental data, even if it takes many years or decades or even centuries before being able to do so. Also, if an agreement between theory and experiment clearly validates a model, discrepancies are used not to invalidate the whole model but only part of the model’s hypotheses. Physics is a so-called hard science but it balances between the hard reality and the rich possibilities of inexact mathematics. The net result of these “arrangements” taken with mathematical power has led to the building of a solid and extended core of understanding of the laws governing inert matter. This is why while discussing the “reality” of Asimov’s psychohistory we emphasize that only a physicist could create it, not a mathematician. Within my sociophysics approach, I am going to follow the same procedure, modeling social reality with an empirical use of mathematics, giving up rigorous aspects, in favor of the capability of being able to reproduce some particular phenomenon. Behind such an apparent lack of rigor lies the basic difference between mathematics and physics. The former focus is on proving the universality of a result, whereas the latter one aims at finding some minimum conditions which are able to reproduce a given observation. However, at the current initial level of our novel approach, we are not yet dealing with what could constitute the associated experimental apparatus of sociophysics. We will first need to elaborate what could be the equivalent elements of the limited experiments performed in psychosociology before even considering larger scale experiments. Dealing with human beings is of course not the same as experimenting with some piece of inert matter. At the same time, this ethical prudence can turn out to be dangerous in the final assessment of a social law derived from our models. This is why we should always be cautious about the findings from a model. At this point, we are not going to state definite laws, but instead, to set out a new framework of investigation in order to be able to gain a coherent and different view about social and political behavior. In a second step in developing our approach, the model predictions have to be checked against real experimental data to validate the coherence of a model’s hypotheses. A whole new framework could thus emerge and gradually be built through the intricate combination of theoretical experiments and real life. It is a long and 28 2 The Question: Do Humans Behave like Atoms? exciting adventure to be engaged in, because so far, we have only just scratched the surface of this new field. What we have done is already a huge and very ambitious epistemological step.
2.4 Implementing a Physics-like Approach Outside Physics
We can now start embracing our main topic, i.e., why physics should be used for tackling anything else outside of inert matter. At this stage, it is worth emphasizing that it is of interest to focus on one particular field, known as statistical physics. Indeed, thanks to its many achievements, the domain has already generated a lot of applications in many fields outside physics. However, the stakes are very different in respect to the applications in biology, economy, sociology, and politics (Fig. 2.8). In a very schematic manner, we can say that in biology there exists no major current interest in theory per se. It is essentially an experimental science. In the case of an experiment being suggested, on the basis of physics-driven ideas, or anything else, it could be performed. If a new result is obtained, the inspiration from physics takes second place and is forgotten. On the contrary, if no result is obtained, the physics is also forgotten.
Statistical Physics
Biology Who cares?
Economy and Finances Why not try?
Social and Political Sciences Very promishing Very dangerous Fig. 2.8 Applications of To be used with care statistical physics to biology, economics, social, and political sciences 2.5 But Indeed, Do Humans Behave Like Atoms? 29
With respect to economics, I would say, in a provocative manner, that nothing really works when confronted to economics data. In finance studies, a scheme derived from physics can be sometimes beneficial to certain types of investment. Many physicists have been hired in financial institutions for their general skills rather than for their use of physics. In particular, a field called “econophysics” has emerged and was developed in the late 1990s with hundreds of scientific papers published, focusing mainly on the analysis of financial data. I am inclined to say that, after serious hopes at the beginning, no solid achievement has been accomplished at all. Of course, as the recent 2008 financial crisis has shown, when inopportune use of abstract models leads to substantial losses, physicists will be blamed for some time to come. But as with biology, the intrusion of physics causes no decisive impact on the dynamics of the field. Contrary to both biology and economics, the use of physics-based ideas, concepts and techniques dealing with human beings may turn out to be very delicate in the social sciences. This statement is particularly true for models which appear to be false. The major associated epistemological contradiction lies in the fact that because physics can contribute substantially to the social sciences, it can also lead to all kinds of serious misuses. The above challenges are related to politics, philosophy, religion, and everything in which human beings are very involved. Convictions, beliefs, and prospects on what is or should be the social life and its organization could make the human cost of both errors and manipulation potentially very high. On this basis, nothing should be imposed in the name of physics. Physics can only shed new light on social phenomena but never substitute itself for the whole framework of the corresponding traditional views and knowledge of sociologists. Hence, sociophysics is a very exciting and promising field of research, but must always be dealt with care and much caution. To sum up the challenge, we could say that “to introduce formal rationality in human beliefs may produce irrationality in human behavior” establishing catastrophic paradigms that are totally artificial and false, although they could fit to some wrong representations of social reality (Fig. 2.9).
2.5 But Indeed, Do Humans Behave Like Atoms?
When dealing with social situations it is usually believed that one of the main difficulties arises from the rich variety of individual characteristics and features of the individuals involved. Along these lines, the richness of a group is expected to be an exponentially increasing function of its size, symbolized by the concept of “complexity” with its famous dogma “The whole is greater than the sum of its parts.” However, crowds, which contain large numbers of people, behave in some aspects like one “collective individual,” who in turn might well behave according to a simpler framework than that of a normal individual. 30 2 The Question: Do Humans Behave like Atoms?
Fig. 2.9 Discovering the existence of deterministic behavior may be the key to individual freedom
This paradox suggests that within a group, the individual complexity should decrease in parallel to the appearance of a new individual monitored by the “collective dimension” of the group. In this context, it is of particular importance to discriminate between, on the one hand, the properties associated with purely individual characteristics, and on the other, those properties which result from the existence of a collectivity or social system. This is exactly why we are not suggesting using, for instance, atomic physics to describe individual properties. On the contrary, collective properties may obey universal features which are valid in many different fields. And it so happens that the interplay between microscopic and macroscopic levels has been greatly studied in physics, a field which is very far from the social sciences. The field of statistical mechanics has been dealing with collective behavior in matter for more than 100 years with much success. However, only in the last few decades has the problem of collective phenomena been well understood. However, this achievement has been made only in the case of pure systems. Organic disorder and heterogeneous systems are still resisting full understanding although solid progress has been accomplished in recent decades with the framing of a series of new concepts and sophisticated tools but much more needs to be done. Out of equilibrium systems are only at the first stages of comprehension. Nevertheless, the modern theory of critical phenomena already represents a substantial qualitative leap toward the understanding of collective phenomena. Chaotic behavior represents another epistemological leap. The modern theory of critical phenomena is based on the fundamental concepts of universality and irrelevant variables. These two concepts mean that different physical systems, for instance, a magnet and a liquid, behave in the same way when passing from one macroscopic state to another, although the physical properties 2.6 Building Up an “AtomÐIndividual” Connection 31 associated with the respective macroscopic states of a liquid and a magnet have nothing in common. Corresponding well-known examples are the magnet becoming a paramagnet and the liquid, a gas. The discovering of the concept of universality has been a major breakthrough in our understanding of the universe. Accordingly, physical characteristics of systems, i.e., the form of microscopic interactions and their physical nature have no effect on the so-called critical behavior which produces the physical character of the transition from one state to another. Most of the microscopic properties turn out to be irrelevant for describing the macroscopic change, which in turn appears to be universal. By irrelevant, I mean that the microscopic properties have no effect on the process. While the number of physical systems undergoing phase transitions is infinite, all associated phase transitions have been shown to be described in terms of a finite number of universality classes. Only a few parameters, such as the space dimensionality, determine which universality class the system belongs to. The abstract and general nature of the statistical physics framework makes it tempting to extend such notions to nonphysical systems, and in particular to social systems for which, in many cases, there exists an interplay between microscopic properties and macroscopic features.
2.6 Building Up an “Atom–Individual” Connection
Nevertheless, the fields of the physical sciences and the social sciences are a priori rather different, both in their nature and in their applications. In order to develop some common framework, it is useful to find out what appropriate assumptions need to be made about the relevant correspondences. The first immediate possibility is to put in parallel the atom and the individual. In physics, the atom defines the basic level of investigation. Afterward, the exploration can go, on the one hand, inside the atom toward elementary particles, and on the other, toward bulk matter by grouping atoms together. In the social sciences, starting from the single human individual also allows us to consider human beings in the “bulk” form, with the existence of societies, as well as the “infra” individual at the level of elementary cells or small groups. Before implementing such an “atomÐindividual” connection, one is entitled to ask the following naive question: “what is common to an atom and a human being?” Without requiring much elaboration, the answer is, “nothing.” The same “nothing” also holds when comparing an atom to a country, a firm, a cell, a political party, a stock, a grain of sand, and many other entities. Such a negative assessment promptly raises a serious concern of what the book is all about. Why all these words for such a clear and precise hopeless conclusion? Without throwing away our nice motivation to build up a new scientific approach to tackle human behavior, we need to go ahead and consider not the singular entity but its plurality. Accordingly, we reformulate the question. Instead of “what is common to an atom and a human being?,” we ask: “what is common to an assembly 32 2 The Question: Do Humans Behave like Atoms?
Fig. 2.10 What is common WWhathat ddoo ttheyhey hhaveave iinn iinn ccommonommon ? to a human being, a stock market, an atom, a party, a grain of sand? A human
A stock 8.2331 An atom
A grain of sand A party
of atoms and an assembly of human beings?.” That already is much more attractive and sounds like deserving of a positive answer. Unfortunately, after some thought and analysis, we need to go against what was expected, with again the same answer of, “nothing.” And again, this remains valid for an assembly of countries, firms, cells, political parties, stocks, sand grains, and many other groups of entities. At this stage, it would be legitimate to get a bit annoyed and to start to complain about what seems to be a totally fake enterprise. However, the above fastidious search was aimed at shedding light on the fallacy of the immediate connections, one atom—one human being and many atoms—many human beings, which have to be dismissed from the start as philosophical traps. Once this is achieved, we can elaborate on climbing out of our despair in pursuing the discussion in order to discover the relevant path to implement the “atomÐ individual” connection. The right and powerful question is: “is there something in common with both processes of passing respectively from one atom to many atoms and from one human being to many human beings?” The answer becomes “yes, there is a lot.” This statement also holds true when going from one country, one firm, one cell, one political party, one stock, one grain, and any other entity toward many countries, firms, cells, political parties, stocks, grains, and other entities. More precisely, the hypothesis behind the present approach is that these microÐ macro mechanisms are universal and hold true beyond the true nature of the various entities involved. The above series of questions aim at clarifying which problems may be addressed and which are outside the scope of the approach. It is worth stressing that we are not claiming that our model will explain all aspects of human behavior. Like any modeling effort, it is appropriate only to a few classes of phenomena of social sciences and not to others (Figs. 2.10Ð2.18). 2.6 Building Up an “AtomÐIndividual” Connection 33
Fig. 2.11 What is common WWhathat ddoo ttheyhey hhaveave iinn ccommonommon ? to a firm, a cell, a country, or whatever?
A firm
A cell
A country
.... A whatever
Fig. 2.12 Unfortunately or fortunately there is nothing in common between all the preceding items
Fig. 2.13 What do a group WhatWhat iiss ttherehere iinn ccommonommon of human beings, stocks and betweenbetween groupsgroups of?of? shares, atoms, and political parties have in common? Humans
Stocks 8.231, 5.217, 9.3
Atoms Parties 34 2 The Question: Do Humans Behave like Atoms?
Fig. 2.14 What do a group WWhathat iiss ttherehere iinn ccommonommon of grains of sand, firms, bbetweenetween ggroupsroups oof?f? countries, and cells have in common?
Sand grains Firms
A firm
A firm
Countries A firm
Cells
.... A whatever
Fig. 2.15 As for previous series, unfortunately or fortunately there is nothing in common between all the preceding items
2.7 Our Bare Methodology
Having presented the focus of our philosophy and the appealing, but wrong paths, we have thus established a solid conviction about the well-founded basis of our approach, at least at an epistemological level. Ultimately, it will be the demonstration of the adequacy of our models in describing some part of the social world which eventually will turn our a priori toward making sense through intuitive statements into something that could resemble a hard science. At least, that is the challenge facing us. Before proceeding with the core of sociophysics, it may be very useful to lay down our general strategy, which is grounded on a single methodology that is used throughout this book for all the social and political phenomena presented. We will start by picking out a single global phenomenon or some social practice occurring in society. This choice could result from our feeling that there exists some hidden paradox. It could also be driven from a legitimate character that no one would ever think of questioning. These choices are ours and many others could have been 2.7 Our Bare Methodology 35
Fig. 2.16 What is in WWhathat isis inin ccommonommon common in going from a inin goinggoing ffromrom oonene human being, a stock market, Humans toto mmany?any? an atom, a political party, a grain of sand to a group of humans, stocks, atoms, political parties, sand grains? 5.234
Stocks 8.231, 5.217, 9.3
Parties
Atoms
Sand grains
Fig. 2.17 What is in WWhathat iiss iinn ccommonommon common in going from a inin ggoingoing fromfrom oneone firm, a cell, a country, or ttoo many?many? whatever to a group of firms, Firms cells, countries, or whatever?
A firm
A firm
A firm Cells
A firm
Countries
.... A whatever made instead. Through these choices, we aim at illustrating the powerful method of sociophysics rather than stating a series of absolute social laws. For instance, we will study in detail the use of bottom-up majority rule voting in hierarchical structures (Chap. 7), and the holding of public debates on issues for which decisions are to be taken that have a large public support (Chap. 10). 36 2 The Question: Do Humans Behave like Atoms?
Fig. 2.18 Contrary to the above previous series, still unfortunately or fortunately there is now a lot in common in going from one to many items
Once we have selected the phenomenon or the social practice of interest, we have to clearly identify the associated paradoxical features. For instance, picking up the study of democratic voting in bottom-up hierarchical structures, we emphasize the surprising stability of top leaderships against growing dissatisfaction from members at the bottom. On this basis, although the phenomenon is usually very complex and involves a large number of different ingredients, we apply a brutal simplification by neglecting most of everything. We reduce the phenomenon to a minimum but always preserve a few basic essential features capable of producing some nontrivial dynamics, keeping in mind the goal of recovering the initial “bare paradox.” In the case of democratic bottom-up voting in hierarchical structures, the outcome is a strong dictatorial stability effect in favor of the current top leadership against an eventual massive bottom rejection. The purpose of the “game” is first to check if the chosen mechanism can reproduce the corresponding selected paradox. For instance, can we stabilize a current leadership against an increasingly huge opposition from the base by a democratic procedure? How can repeating votes end up by democratically choosing to renew the current leadership although it is now rejected by the majority of people at the bottom of the hierarchy? Once the toy model is working, the second goal is to establish a logical link between the artificial model and what could be the equivalent real mechanism underlying the political life of institutions and their failure to take into account at the top leadership the eventual changes that have occurred at the bottom of the organization. Somehow, the purpose of the modeling is to create a virtual world from some basic interactions and to study all the associated properties. Once this elementary world is understood and its parameters controlled, we move on by pushing up its basic limits to embody more ingredients so as to make the overall model a bit more realistic. We also need to check the robustness of the mechanism under small changes of the model parameters since we are looking for universal features. In the foundations of sociophysics stands the a priori claim that there exists some finite number of universal mechanisms behind the many different sorts of social behavior occurring in human societies. The study ends by outlining possible extensions of the range of applications of the model. 2.8 To Sum up 37
2.8 To Sum up
As stated in the preceding sections, we will always start by considering one salient paradoxical feature of a particular phenomenon which occurs in society. Although, it is often the result of a very complex situation, we will considerably reduce it down to its simplest form by neglecting many other ingredients, even when they are quite clearly active in the making of the feature. The goal is to undress the selected feature in order to extract some well-defined “pure paradox.” Indeed, the a priori claim that there exist a few universal mechanisms behind much social behavior occurring in human societies is at the very foundation of sociophysics. On this basis, we set up the simplest possible model which can reproduce the selected social paradox using bare and precise mechanisms. The purpose of the “game” is first to check if this bare mechanism can underline the onset of the paradox. Once that is established, the second step seeks to establish a logical link to the supposed feature, in other words, to the real social or political property we started from. Nevertheless, it is worth stressing that it may well happen that the model will be eventually more relevant to another situation than the one from which it was initiated. Once everything in the dynamics of the model is controlled and understood, we move on with the possibility of pushing its basic limits, in order to embody more realistic ingredients. It is also of importance to check the robustness of the result with respect to the chosen bare mechanism, since in particular for social situations, the initial conditions are never exactly known. To grasp the phenomenon requires obtaining results which survive against small changes in the parameter settings. At this level of our investigation, we do not address the question of the quantitative applications to the studied social phenomenon. We aim at defining some qualitative tendencies, which are hopefully present in real-life phenomena, but embodied and masked by some others. Although our models are quantitative in the sense that, using physical modeling and mathematical calculations, they yield precise numbers and figures, these numbers should not be taken too seriously by themselves. Only the associated qualitative description of the considered phe- nomenon is to be given serious thought to eventually deal with social realities. It is on this basis that we will never take too seriously the figures we get but on the contrary, we will focus seriously on the nature and the trends of the various dynamics obtained. The overall descriptions will be put forward. It is also of central importance to stress the following observations. Firstly, the use of modeling in physics has been tremendously powerful in establishing the field as an exact hard science, capable of building concrete and efficient devices. Secondly, this power comes from the empirical uses of mathematics to describe real phenomena. It means that it is not the mathematical rigor which prevails but the capability of reproducing some particular property. Thirdly, and this point is essential, all the results obtained from the various models are aimed, sooner or later, at being tested against experimental data. Fourthly, discrepancies and agreements are used to validate or invalidate part or all of the model’s hypotheses. It is exactly 38 2 The Question: Do Humans Behave like Atoms? this backward and forward motion between theoretical and experimental research that has yielded an extended core of understanding for the laws governing inert matter. Here, within our approach, we are modeling and empirically using mathematics to only proceed up to the first two points. We give up the rigorous demonstration of the model to favor its capability of reproducing some particular phenomenon. While mathematics focuses on proving the universality of a fact, physics aims at finding some minimum conditions, which are able to reproduce the fact. At the present level of our novel approach, we put aside any experimental apparatus. This fact can be dangerous in the final assessment of a social law derived from our models. This is why one should always be cautious about our findings. At the moment, we are not going to state real and definite laws, but instead to set out a new framework for developing a coherent and different view about social and political behavior. This is why our findings are always and systematically to be taken with caution. The ongoing testing of a model’s predictions of real experimental data to validate the model’s hypotheses has to be left to the future whose proximity will be a function of our collective progress on the above level of research. In turn, an equivalence to the whole framework of physical laws that has to be constructed through the intricate combination of theoretical and experimental research has yet to be developed for sociophysics. This fact does not make the project any less ambitious. Last, but not least, all of our modeling is based on both “naive” assumptions of open minded people, with everyone having the same individual power and the existence of open and democratic social spaces. Clearly, although both these assumptions are false, they do not prevent an operative scheme from discovering the secrets of organized human behavior. Accounting for all the complexity of human differences is left to other fields of investigation of human behavior. In short, our general procedure is articulated around the following steps: 1. Choose a particular phenomenon 2. Single out one salient paradoxical feature 3. Define the phenomenon quantitatively down to its simplest form 4. Calculate the inherent “paradox” 5. Find a logical link to a real-life counterpart 6. Push the limits of the model 7. Enumerate possible extensions Our general guide is illustrated in Fig. 2.19. At the core of our fundamental strategy stands the somewhat provocative hypothesis that the conditions for sociophysics are: “Although you may have to accept drastic simplifications and crude hypotheses in order to start a wrong model, later, this could develop into an operative tool with predictions, which in turn may have a good chance of being true.” We can add our founding statement in Fig. 2.20. References 39
Fig. 2.19 Our constant methodology in seven distinct steps
Fig. 2.20 Our provocative, yet very constructive assessment
References
1. S. Galam, Y. Gefen and Y. Shapir, Sociophysics: A mean behavior model for the process of strike, Journal of Mathematical Sociology 9, 1 (1982) 2. S. Galam, When humans interact like atoms, Understanding group behavior, Vol. I, Chap. 12, 293-312, Davis and Witte, Eds., Lawrence Erlbaum Ass., New Jersey (1996) Chapter 3 Sociophysics: The Origins
The origins of Sociophysics are discussed from personal testimony. I trace back its history to the late 1970s. The first steps of my 30 years of activities and research to establish and promote the field are reviewed. In particular, the decades long strong opposition from the physics community is emphasized, together with the almost total absence of contributors. During these years of rejection, only a very few and scarce papers were published in the field. However, in the mid-1990s, quite suddenly the number of contributions started to increase. Up to the beginning of the 1990s, any attempt to apply physics to social sciences was considered as a devaluing of physics, the noble science, “par excellence.” Today that is no longer the case; on the contrary, it is considered as a real challenge for shedding new light on social and political behavior through the prism of physics. The irony of this reversal in process is that the few authors who wrote about the origins of sociophysics had a tendency to set up “nice stories” about it. By the waving of a magic wand, the contentious aspects of the adventure of sociophysics, the one that really took place, are removed completely. Instead, an ideal version of a smooth continuity from social sciences to sociophysics is set up. What happened is reminiscent to me of the way communist regimes used to write history. Maybe indeed it is the way that most of the writing of history is implemented. While in the communist case the falsification was deliberate, for other cases, it could be the unique solution when prominent figures are no longer alive to testify. Moreover, even though still alive, these fighters for knowledge, once they have been accepted and reintegrated within the institutions they were fighting with, want to forget and erase the traces that show that they used to be heretical and marginal.
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 41 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 3, © Springer Science+Business Media, LLC 2012 42 3 Sociophysics: The Origins
3.1 The First Days
3.1.1 Breaking the Secret of Critical Phenomena
To apprehend how sociophysics has emerged from the minds of a handful of physicists, it is of importance to recall what the general atmosphere was at that time. One of the most difficult puzzles of condensed matter physics has been for centuries the understanding of the experimental fact that many physical systems can all of a sudden, at a precise value of some external parameter, such as temperature, undergo a drastic change in their macroscopic state. It is as though the material suddenly metamorphoses itself into another one, like in a chemical reaction or a “magic” transmutation. The problem is that the microscopic nature of the material remains unchanged, indicating that it is not a chemical reaction but a macroscopic reorganization. Moreover, these dramatic collective transformations are accompanied by the disap- pearance of certain macroscopic properties and the appearance of novel ones. The transitions are from one collective order to another, or to a partial or total disorder (Fig. 3.1). The enigma of the so-called critical phenomena resisted all attempts to reveal its secret. A theoretical explanation was missing, although many industrial applications existed. The challenge remained. But in the mid 1970s, a revolution occurred that was driven by the decisive work of Kenneth G. Wilson. He eventually came up with a coherent explanation in terms of universality classes and order parameters. Everything was solved using the powerful techniques of the renormalization group. He provided the magic key to the fantastic worlds of critical phenomena. A good review can be found in [1]. In a similar way to a phase transition, our theoretical understanding of the properties of inert matter jumped from a kind of disordered knowledge to an ordered understanding.
Fig. 3.1 Physicists were puzzled by all the experimental results obtained from materials undergoing a phase transition 3.1 The First Days 43
Fig. 3.2 The puzzle was solved by Kenneth G. Wilson and implemented by Wilson and Fisher. Physicists were looking at phase transitions with great clarity and a non hidden satisfaction
However, Wilson’s initial work solved the problem in principle but was not yet directly applicable to real systems. Shortly after his milestone publications in collaboration with Michael E. Fisher, he devised a scheme under the name of the “epsilon expansion,” which allowed the reaching of the real world that sits in three dimensions [2].
3.1.2 The Physicist’s Corner
Their paper title, “Critical Exponents in 3.99 Dimensions” was deliberately provocative... at that time. Dimensions were by nature integers. Mentioning a rational number for a space dimension was akin to heresy. And indeed it was. Here we have a nice illustration of how physics proceeds using “iconoclastic” statements. The Wilson renormalization group scheme did not allow the making of explicit calculations for systems in three dimensions, i.e., real physical systems. However, since the equations become much simpler in four dimensions, Wilson and Fisher had the idea of making an expansion around four dimensions as a function of the small variable epsilon defined by d D 4 (Fig. 3.2). From a physical point of view, this idea was nonsense, not making an expansion a function of a small variable but solving equations for a noninteger value of the space dimension. Very typical of the physicist’s methodology, once the equations were solved as a function of epsilon, they plugged in the value D 1 to claim that the associated results corresponded to the real space dimension d D 3.Anditworks. This solution meant that the results now fitted to experimental measurements. 44 3 Sociophysics: The Origins
Fig. 3.3 One of the biggest challenges of physics was just solved in an incredibly elegant and powerful framework. The world of physics was blossoming
What became more intriguing is that, as so often happens in physics, some years later, the above trick to overcome a mathematical barrier was given solid validation. Noninteger dimensions were found to exist in the real world. They are called fractal dimensions [3]. These crucial and brilliant steps led to the so-called Modern theory of phase transitions, which propelled condensed matter physics into its Golden Age. Both Wilson and Fisher were professors at Cornell University in America which became the “Silicon Valley” of this new fascinating “gold rush” of Physics. Several universities all over the world promptly created an active network to establish a very active and successful new field of research. Young physicists were entering the field with a great deal of excitement and enthusiasm. Being very productive, they published papers by the hundred. Notwithstanding, although I was one of them, I did not publish papers in such large quantities (Fig. 3.3).
3.2 The First Days, the First Fight
For a series of personal reasons, I arrived in Israel in 1976 after graduating from the Pierre and Marie Curie University (UPMC) in Paris in 1975 with a doctoral degree in physics. Being very excited about the blossoming field of critical phenomena, I embraced the new field by engaging in a second doctoral degree with a Ph.D. on the study of disordered systems at the department of physics and astronomy in Tel- Aviv University (TAU). It was one of the world’s “hot” centers in the field thanks 3.2 The First Days, the First Fight 45
Fig. 3.4 For everyone, to talk about the future frustration of successful physicists was simply absurd to a group of young and smart professors working hard for posterity, as well as for their own reputation. A quite impressive number of solid contributions to the field of critical phenomena and disordered systems were produced there during this period. Already having in mind the project of building sociophysics, my view was to promote a collective project. At that time, socialist ideas were all the rage in the Western world and many dreams were attached to them. Along the associated dialectical frame of mind which prevailed then, I foresaw a growing contradiction between on the one hand, the power of the concepts and tools of statistical physics, and on the other, the imminent exhaustion of possible subjects of interest to which they could be applied. From that idea, I really questioned the future of the role of physicists in the development of knowledge at that time (Fig. 3.4). My idea was that as physics would reach the limits in understanding of inert matter worth investigating, a frustration among physicists would appear. To compensate for their increasing frustration combined with their agility in using their particular methods, they would start applying physics outside of the field of physics. A kind of imperialism among physicists would lead to the invasion of the various other fields of research in science. Today the invasion is at its paroxysm as can be seen, for instance, from the incredible development of biomaterials, nanotechnologies, biomagnetism, etc. Within this frame of thinking, I was suggesting that physicists should deal with the rich variety of behavior related to human activities, including economics, sociology, psychology, and politics. 46 3 Sociophysics: The Origins
Fig. 3.5 Physics may provide some useful tools and concepts to tackle the twenty-first century. I advocated that program at Tel-Aviv University (TAU)
I published a series of papers to justify and elaborate my suggested strategy to escape the physicist’s frustration, several coauthored with Pfeuty [4Ð10]. The focus was on the existence of some epistemological contradictions within physics and among physicists. Unfortunately, I was a bit naive and almost every physicist who read my work, leading and nonleading, young and old, strongly rejected my proposals. Daring to compare human beings with atoms was strongly condemned as a blasphemy to the pure sciences and an insult to human complexity. It was looked upon as being total nonsense, something to be rejected. And it was indeed fully rejected for more than 15 years (Fig. 3.5).
3.3 From Claim to Demonstration
While going around like a self-convinced salesman trying to sell my idea of what should be done to avoid physicist’s frustration and to enrich the scientific understanding of human behavior, I found out about two connected papers in the literature. I was rather happy to discover that some other physicists had already engaged in my program. The first paper I found was from 1974 by Callen and Shapiro. They wrote a short note in Physics Today to suggest a promising similarity between fish band imitation and Ising spins [11]. It was just a nice metaphor, which, to my knowledge, unfortunately had no follow-up. Three years earlier, a much more elaborate paper was published by Weidlich in the British Journal of Mathematical Psychology [12]. It dealt with the dynamics of opinion forming. Although more developed and dealing directly with a social phenomenon, the paper’s content was more of an application of partial differential equations in the spirit of Volterra’s work than an application of the Modern theory of phase transitions (Fig. 3.6). 3.3 From Claim to Demonstration 47
Fig. 3.6 The first fight was to advocate sociophysics
Then, while continuing to argue with physicists about the possibility of socio- physics I wrote my first contribution in 1982 [13]. Disturbed by the philosophical statement of an eventual thermal death for humanity, I took an opposite stand, arguing that applying the Carnot principle of maximum entropy to societies was a great perspective of freedom and development for mankind. Contrary to the asserted ineluctability of the ultimate thermal death of human societies, there are plenty of reasons to become optimistic about them. Nevertheless, it was of a factual contribution written in French and published in a philosophical journal. It is not surprising that it had only a virtual existence in terms of impact. Some time later, with two Ph.D. colleagues from the same department of Physics and Astronomy at TAU, Yuval Gefen and Yonathan Shapir, I published a paper to set simultaneously a global framework for Sociophysics as a new field of research and to make a first application linking the phase transition of an Ising model to the process of strikes in an industrial plant or a company [14]. The stable state of the agents, working or striking, was studied in particular as a function of an external reversing uniform field competing with nearest neighbor interacting agents. The rather subtle phenomenon of metastability combined with the hysteresis feature were used to give political and social explanations. Besides its scientific content, the paper contains a call to the creation of sociophysics. It is a manifesto about its goals, its limits, and its dangers. As such, it is the founding paper of Sociophysics [14], although it is not the first contribution per se (Fig. 3.7). We chose the Journal of Mathematical Sociology to submit our manuscript and not to a physical journal since I first thought sociophysics should be naturally published in a social sciences journal, although it had been developed by physicists. Later I changed my mind, focusing mainly on physical journals for my subsequent publications. It took over 2 years to get our paper finally accepted for publication after a series of tough exchanges with several referees. The paper appeared in 1982. 48 3 Sociophysics: The Origins
Fig. 3.7 The manifesto of sociophysics. A paper by Galam, Gefen and Shapir published in the Journal of Mathematical Sociology 9, 1 (1982)
All three of us were pursuing our physicists’ careers and went to the United States as postdoctoral students. We did not hear much about the earlier fate of the paper. It is worth stressing that at that time there were neither internet nor emails, nor cheap international telephone calls.
3.4 The Story Behind the Scene
I will now tell the story of the sociophysics manifesto, the one that otherwise would never have been written and thus never known. This story puts all the politically correct writing of the sociophysics history in big trouble. Here below is what happened at the Department of Physics and Astronomy at TAU. In these earlier “underdeveloped times,” when you wrote a research paper, you needed first to write it by hand, with all the equations, numbers and references and without forgetting to draw the pictures, the most tedious and delicate part of the task. Once everything was completed, you gave the manuscript to the secretary in charge of typing it. At the time it was a highly specialized skill which once started, took several days. One has to imagine how rudimentary and slow the process was to type up in Latin script, plus putting in the integral signs, and so on and so forth. It was almost like a dentist operating in an open mouth. 3.4 The Story Behind the Scene 49
Fig. 3.8 In a world level university, at the end of the twentieth century an attempt is launched to muzzle a researcher and to deprive him of the right to submit a manuscript
Once the paper was eventually typed, a careful check was in order but only to correct mistakes. It was out of the question to rewrite the whole paper or part of it as everyone does nowadays, thanks to computers, great software, and Latex (Fig. 3.8). After having gone through such an amazing challenge, we were very anxious to get the typed manuscript, the first incarnation of the printed material before submitting the paper. Moreover, it was for each of us, more or less the first “real paper.” Running and jumping, we raced to the secretary to take delivery of our document but the unthinkable, a fair “Coup de the«atre,”ˆ happened. Our paper had been taken away by the Head of the Department and locked in a safe cupboard! A scientific paper written by three Ph.D. students had been literally sequestrated without notice under the authority of the Head of the Department and moreover, with the support of most of the faculty members. We were totally astounded. How could such a thing happen? The “event” provoked a big scandal within the department. We were denounced as putting at stake the department’s reputation of excellence. While we were claiming our right to academic freedom the chairman denied it using the fallacious argument that we had no tenure positions. Our manuscript was simply under arrest (Fig. 3.9). Trying to find a way out, it was settled that to satisfy academic freedom, non tenure fellows must obtain the endorsement by a tenure professor of the department. In an explosive atmosphere, a former member of the refusenik seminar from the Soviet Union, a well known experimental physicist and then full professor at TAU, came to our rescue. Alexander Voronel stood to grant us the freedom to recover our manuscript and, moreover, to submit it to an international journal. And so we did [15](Fig.3.10). 50 3 Sociophysics: The Origins
Fig. 3.9 In a world top university, at the end of the twentieth century, a manuscript had been confiscated without notice by the Head of the Department of physics with the support of most of the faculty members
Fig. 3.10 The muzzled researcher struggled for his freedom, and in the same department a courageous professor from the Soviet Union stood up successfully to ensure that full academic freedom was granted 3.5 More About Academic Freedom 51
To ensure academic freedom does not guarantee the publication of a manuscript. It only imposes the freedom to submit a manuscript, which then has to be refereed by peer experts from the journal to which it is submitted. This is the way that research publications have to operate. Apart from it being interesting in itself, the above story illustrates how much opposition there was to sociophysics from within physics itself, i.e., from physicists. And that was not specific to the physicists of TAU; the hostility was felt from most physicists around the world and I had the chance to meet a good number of them. Indeed, I could tell many stories of this kind, which occurred to me during these many years of a personal and lonely fight to create, develop and establish sociophysics. The opposition always came from within the general framework in which I was positioning my various contributions. It was the idea of creating a new field of research within physics in order to deal with human behavior which was deeply disturbing the physicists, not a factual contribution which could always be looked upon as a marginal and exotic isolated event with no follow up.
3.5 More About Academic Freedom
From these later years of the twentieth century in the late 1970s at the department of Physics and Astronomy at TAU, I accomplished a long and well-settled career. I held successive positions at different academic institutions including, in chronological order, the City College of the City University of New York (CUNY), the New York University (NYU), the UPMC in Paris, and the Ecole« Polytechnique. In 1984, I obtained a permanent position at the CNRS, the National Center for Scientific Research in France where I am currently a Director of Research, the equivalent of a full professor position but without teaching duties. My career was built upon my research accomplishments within traditional physics, although on a slowed down path due to my “other” activities. Not withstanding the negative view of both institutions and individuals for 20 years, sociophysics became a recognized and flourishing field of physics that involves today a large number of physicists all around the world. Amazingly, in the early twenty-first century I again went through a very similar experience of near ostracism from colleagues who feared for their reputation. In February 2007, I wrote a paper in a major French daily newspaper, Le Monde, in which I questioned the claimed human culpability with respect to the assessed global warming [16]. I stated that there existed no scientific proof of the Intergovernmental Panel on Climate Change (IPCC) climatologists’ claims. Taking such a position at the time resulted in quite an outcry against me and the newspaper, Le Monde, which had never experienced such a shaking from both outside and inside the newspaper [17]. I later wrote a book to detail my arguments but this is outside the scope of the present book [18]. 52 3 Sociophysics: The Origins
Fig. 3.11 A quasi unanimity of physicists looking at sociophysics during both the 1980s and the 1990s
The above cases demonstrate that modern man has not changed much and still relies on archaic and primitive reactions as soon as he fears that something could jeopardize his status. It applies to scientists in the same way as to anybody else, which is a good lesson to keep in mind. Novelty is still perceived as being dangerous even by those who have even contributed to it in the first place (Fig. 3.11). I have nevertheless got to put some caution on my plea for academic freedom. I am aware that it could open the path to the possibility for a faculty member to write “unacceptable” statements, in particular in the various fields of social sciences. But in this case, it should be the civil law that should be activated to eventually condemn a published paper that violates some existing laws. But it is not the role of the faculty to enforce an a priori censorship on any submitted publication. For sure, if a faculty member is prosecuted and proved guilty, for instance of racist incitement, or any other criminal act, to have them fired by the faculty makes sense. Such a statement is of course valid as long as we are dealing with a democratic country.
3.6 Surviving Within Physics by Not Playing Tennis
I was totally convinced that sociophysics was about to be established in the future but I was too much ahead of my time. Therefore, in order not to put at stake my ambitious program it was clear for me that I had to stay within both the academic world and the physics community. Only there could I eventually succeed in my project of creating a new field for the social sciences. 3.6 Surviving Within Physics by Not Playing Tennis 53
Fig. 3.12 To survive and to become better equipped for the adventure of the building of sociophysics I had to do solid research in orthodox physics. At the same time I was presenting my sociophysics activities as a hobby. Others were playing tennis, while I was “playing” sociophysics
Accordingly, and so as not to jeopardize my academic career, I had to divert most of my energy away from doing full time research in developing sociophysics. The goal was not only survival as a physicist but also to blossom within physics. I thus did orthodox physics, which I must say was at that time a really exciting activity. It embodied the then new field of disordered systems from which I was sure that many ideas, concepts and tools could be used later on to produce sociophysics models. I also carried on working a bit on sociophysics, not in secret, but as a “hobby.” In the way others were playing tennis, I was “playing” sociophysics. It provided me with some kind of immunity against any criticism of lack of depth from scientific evaluating committees. As a matter of fact I did survive as a physicist. I even accomplished a fair career. During these years, physics was highly competitive and not too many openings were available around the world. The price of such an achievement was a heavy personal involvement in research to allow success in physics as well as novel work in sociophysics. It was only a few years later, while in New York, that I was able to produce more work to further establish the feasibility of Sociophysics (Fig. 3.12). I first published an additional paper using entropy in 1984 [19]. It was still about concepts and ideas but without a quantitative construction. My second significant contribution to sociophysics appeared in 1986. Using the fascinating power of renormalization group concepts and techniques, I studied dictatorship effects induced by the use of the democratic rule of majority voting in hierarchical bottom- up organizations [20]. Still thinking that sociophysics should be implemented within the social sciences, I submitted my paper to the Journal of Mathematical Psychology. The same scenario as with the manifesto paper occurred [14]. Two 54 3 Sociophysics: The Origins
Fig. 3.13 Still trying to convince others to engage in sociophysics. Zero success
long years of ongoing arguments were necessary with several referees and one of the journal Editors in order to have the paper accepted. Once the paper was printed, I did not get many reactions. But still being very enthusiastic and well orientated on my particular path, I tried to create a collective movement rather than to build an individual effort. I thus organized several informal seminars and kept on trying to convince other researchers of the validity of the approach (Fig. 3.13). After several years in a row without going back to Israel, I went to TAU for a short visit. At the same time, Dietrich Stauffer, an expert physicist in Monte Carlo simulations from Koln University in Germany was also visiting the department of physics and astronomy. We shared the same office. He was then doing simulations related to physics but became interested in my approach to sociophysics. He encouraged me to submit a sociophysics paper in a physics journal. Back in France, I followed Stauffer’s suggestion and submitted a paper on voting to the Journal of Statistical Physics, which looked like a good choice to try and publish a first publication of sociophysics for physicists. It is worth noting that although the paper was accepted in 1990, I received quite a surprising letter form the chief Editor Joel Lebowitz from Rutgers University. He made a comment on his own decision to accept my paper for publication [21]. He stressed explicitly that he was accepting the paper because the referees’ reports were positive and that the associated editor in charge of the paper recommended publication. However, he personally did not believe at all in the validity of such an approach although he did not intend to censor it. He thus respected and implemented academic freedom against his own feelings. Such an exemplary attitude has to go to Lebowitz’s credit. But once again, I did not receive much feedback after publication. Making the most of my still active enthusiasm I later published three more papers to extend my voting model [22Ð24]. For each of them, the same long and involved process ended up with no reaction to their publication. 3.7 Breaking the Gap with a Social Scientist 55
Fig. 3.14 In the late 1990s at last the hostility against sociophysics faded
This funny business went on for more than 20 years until my peculiar concept of “not playing tennis” became popular among a substantial number of other physicists around the world. Sociophysics was thus established de facto not by institutional decisions but somehow against the institution’s will, in the field of research itself. It has been the very existence of sociophysics throughout the first several dozen and then the hundreds of research papers published in leading international physics journals that has in turn validated the field (Fig. 3.14). To bring a happy ending to the story, I must mention that it was finally, in 2004, that my institution acknowledged that my unusual practice of “tennis” was not a hobby but in fact my professional activity. In other words, I was officially granted the right to carry out full time research in sociophysics. I implemented this change in status by quitting my physics laboratory to join a social sciences group at the Ecole« Polytechnique where I have been ever since, as happy as a fish in water. I must put in a word of caution, to stress that although I have been formally allowed to embrace sociophysics, the counterpart has been an end to my institutional career in terms of an eventual promotion (to Director of Research, first class). But one cannot ask too much from a national institution. And more importantly, when you engage in new spaces to break the existing frontiers, whatever their respective nature, you do not expect a straightforward career. You just hope to survive and keep on thriving on your own intuitions.
3.7 Breaking the Gap with a Social Scientist
Having taken so much effort and made so many contributions with no visible reaction from either physicists or social scientists, I thought of changing my strategy. 56 3 Sociophysics: The Origins
Fig. 3.15 An improbable meeting between two French Serge, one a psychosociologist and the other, a physicist, in Manhattan
The correct next step was maybe to take with me my new field of sociophysics and to seek to collaborate directly with a social scientist in studying a problem that social scientists had tried to solve without any success (Fig. 3.15). Instead of just publishing lonely papers in social science journals, elaborating joint work with a social scientist could trigger a drastic change in both the fields of physics and the social sciences. Researchers are often very naive. At least I am. Having decided this reorientation, everything still had to be done. I had to find someone who would not only be interested in the project, and who would not only be ready to invest his or her time (no money was required), and would not only have to do this and that, but who would also be ready to invest tremendous patience in the willingness to be at the same time both a teacher and a student, in addition to being engaged in subjects for which sociophysics could be helpful. As so often happens for improbable tasks, things moved on quite by chance. I was in New York enjoying the Big Apple way of life of drinking a good expresso in the West Village when I saw a local friend walking with somebody. Joining me for a drink it happened that the other person was Serge Moscovici, a leading French social psychologist, who was visiting New York. I immediately put forward my case. He appeared rather interested in the adventure and we decided to meet again once back in Paris. We got along well and started a very fruitful cooperation which lasted over a few years from 1991 until 1995. We came up with a novel theory of decision making, which yielded a series of papers, most of them published in the European Journal of Social Psychology [25Ð29](Fig.3.16). But again, a lot of work, a lot of brainstorming, a lot of time, a lot of excitement, a lot of effort, a lot of rewriting, a lot of arguing, a lot of... to end up with not much 3.8 Changing My Strategy: Back to the World of Physics 57
Fig. 3.16 A very fruitful collaboration created a novel theory of group decision making embed- ding both sociophysics and psycho-sociology feedback, except one invited paper in a book edited by a social scientist [30]. The papers had too many equations for social scientists and too much psycho-sociology for physicists, who anyway ignored the existence of this peer reviewed international journal. Often, people discussing interdisciplinary research point to the necessity of having researchers from different disciplines working together. They always talk about the pay-off of such collaboration, pointing to the resulting benefit for the research. Unfortunately, my few experiences of real interdisciplinary work indicate that one should not mention pay-off but more the heavy price to pay for those doing the research. Of course, I am not discussing here the many “fake” collaborations where people from different disciplines coauthor a paper, which is in fact monodisciplinary and written only by the experts of that discipline. In such a case, everyone receives a return from their own community demonstrating their interdisciplinary involvement by having a coauthor from another discipline.
3.8 Changing My Strategy: Back to the World of Physics
Having first published my papers in social science journals, I then sought to create a solid interdisciplinary theory with a social scientist. I eventually came to the conclusion that this was not yet the optimum way of going about things. Coming back to my initial postulate about sociophysics emerging from physicists, I realized that the natural logic of such a statement was to publish sociophysics papers in physics journals and not in social science journals. This is what I started to do [31, 32] in the mid 1990s (Fig. 3.17). 58 3 Sociophysics: The Origins
Fig. 3.17 Econophysics jumped onto the stage. Sociophysics was still in the shadow of loneliness
However, while I was engaged on this new path, I stayed just as lonely as before. There were not many reactions, apart from a few hostile ones from time to time. The main reaction was just an absence of reaction, a total indifference. Finally, around the same time a handful of physicists also turned “exotic,” not in joining sociophysics, but in analyzing financial data and stock markets. It was the birth of the so-called “econophysics” as coined by H. Eugene Stanley from Boston University in the United States at the conference “Dynamics of Complex Systems” held in Kolkata in 1995 [33]. It was the first validation of my prediction I made in the 1980s of physicists invading new grounds of investigation outside the scope of inert matter. Although it was a satisfying observation, it did not put an end to my loneliness in my obstinate construction of sociophysics to deal with the study of political and social behavior.
3.9 The Secret One Shot International Seminar
3.10 The Rising Sun of Sociophysics
Years were passing by, one after the other, which was the minimum requirement for maintaining my obstinate fight in establishing sociophysics. At around the twentieth of them, all of a sudden a few papers appeared in the literature embracing the sociophysics approach. Several physicists here and there joined the club by making sporadic contributions. I was so happy to realize that I was not really crazy, or at least not the only one who was so. My dream was becoming a real and solid reality (Fig. 3.18). 3.10 The Rising Sun of Sociophysics 59
Fig. 3.18 Finally, a few physicists here and there started to publish sporadic work embracing my sociophysics approach. I was so happy at last
However, my happiness turned rather too quickly into frustration. Most of these new papers did not cite my earlier founding papers. Even worse, some were “rediscovering” my own contributions and appropriating the paternity to themselves. The emerging young community was endorsing this incredible on- going practice. Sociophysics was blossoming almost as if neither I nor my earlier papers existed. What a disappointment! My naive belief that physicists were more honest than most others when dealing with science was shattered. The very human world of science was just as dishonest and double-faced as any other social sphere. In fact, such a phenomenon is rather common in science. The only difference with the other worlds is that science pretends to be honest at its foundation in order to ensure its progress. But of course, as usual, there exists a gap between the idyllic representation of reality and reality itself, often a big gap (Fig. 3.19). I thus began a new unexpected struggle to try to restore the truth about the paternity and chronology of the development of sociophysics. But it became rather quickly an overwhelming task due to the success of this new field which was expanding at a fast rate. Therefore, from time to time I sent occasional e-mails to the relevant authors mentioning my previous contributions. But overall, the outcome was that I felt rather at unease. I felt so much of an incredible injustice. The alternatives were and still are either to witness with frustration certain colleagues in this new field being promoted for supposedly “original” work that I in fact did many years ago or to appear as a frustrated paranoiac. The choice is hard to make. Indeed, I have oscillated between the two options at each new occurrence of this spoliation. It is unpleasant and psychologically consuming to play the role of a bitter scientist when you just want to restore the chronology that really happened against the one which is wrongly stated by powerful scientists who are building the institutional frame for the field (Fig. 3.20). 60 3 Sociophysics: The Origins
Fig. 3.19 The new sociophysicists were pretending to be the first ones, ignoring my series of earlier pioneer contributions to the field. My happiness turned into a feeling of injustice
Fig. 3.20 I found myself engaged in a new and totally paradoxical struggle for recognition by the new sociophysicists
3.11 When Too Much Is Too much
To top all of this manipulation and dishonesty is the writing of a history that denies the real history. Writing a past history is a difficult task due to the necessity of rebuilding a world that no longer exists by only using the often incomplete documents that are available. But writing a contemporary history is of a different 3.12 Claiming the Paternity of Sociophysics 61 nature, since usually some of the actors of the event are still alive and can testify alongside the associated data and documents. However, I have been astonished and shocked when I read some of the first papers about the history of sociophysics. They totally ignore the real history, the one which actually happened. To illustrate my case, I will cite two papers discussing the nature and origins of sociophysics [34,35]. I am not discussing this point just to restore my “battered” ego, but mainly to settle the historical truth on the origins of sociophysics. To strengthen sociophysics as a solid field of research it is a prerequisite condition to preserve its conflicting nature. It is an essential ingredient of the approach. Both of the above papers are different in their focus and style, but both adopt what could be defined as a very politically correct view of the dynamics of science. They introduce sociophysics as a natural outgrowth of sociology, and connect a series of papers that in particular ignore my own. On this basis, they build a theoretical history driven by a certain logical link among these selected papers. This “creative history” may be appealing intellectually but it is quite simply false, particularly with respect to both the scientists involved and the epistemological content. To support their idyllic view of the origins of sociophysics, they trace back its foundations to the work of Schelling [36] who according to them was already doing physics without being aware of it. It is not my intention to minimize the substantial contribution of Schelling to the social sciences but his work is not sociophysics and did not contribute to the first steps of the emergence of sociophysics. A few years after sociophysics began to spread, Schelling’s work did lead to a series of papers by a few “newborn” sociophysicists. Nothing less but nothing more.
3.12 Claiming the Paternity of Sociophysics
To claim a paternity is rather awkward in particular with respect to ideas. The educated and humble approach is to let others attribute a paternity to someone. It looks arrogant and inappropriate to do so oneself. However, my approach relies on the idea that reality must prevail against a hypothetical “savoir vivre.” At some point of observation of what was going on, I considered the necessity to claim the paternity of sociophysics (Fig. 3.21). By so doing, I am fully aware that to be a “father” does not mean to be the first one ever. One can always find in the past here and there some figures who evoked the idea or even created some local metaphor, which nevertheless did not have any follow up. It is similar with making love and having a baby; the father of a child is not necessarily the first man who had intercourse with the mother. I acknowledged this fact explicitly by giving references to the few papers I knew of, which were related to sociophysics in one manner or another [11, 12], in the paper in which I put forward as the manifesto of sociophysics [14]. It is also of vital importance to avoid the misuse of imported concepts from physics, as was well demonstrated by A. Sokal and J. Bricmont in their book entitled “Intellectual Impostures” [37]. 62 3 Sociophysics: The Origins
Fig. 3.21 I never thought about paternity or recognition during my more than 20 years of fighting. But then, facing the threat of being usurpated, I decided to claim my paternity
For 20 years I was involved in promoting sociophysics, fighting without having in mind the question of paternity of the field. The awareness of being the father of sociophysics came along one evening in 2000 while drinking a beer at an international conference at Bad-Honnef in Germany with H. Eugene Stanley, a prominent world wide American physicist from Boston University. He knew me from the 1970s on my arrival in Israel and had witnessed the whole of the period up until now. The conference was entitled “International Workshop on Economic Dynamics from the Physics Point of View” and my talk’s title was “Random field Ising model for group decision making.” While discussing the new reality of physics going beyond inert matter Gene Stanley made the statement that indeed I was the father of sociophysics. This was like a shock for me since the concept of fatherhood was so far from my preoccupations. I was motivated by “love,” not by having a family! I thus kept this “revelation” hidden in my mind. Later on, with hindsight I came to the same observation: Gene Stanley was reporting a fact. But yet, I kept this fact “secret” up until another international conference in 2003 where I was just “pissed off” by a talk in Poland about sociophysics given by Dietrich Stauffer. There, my contributions were totally ignored. The day after, without notifying the organizers, I took a few minutes of my invited talk to put things straight: I was the father of sociophysics. Later, I published a paper [15] to provide my claim with more details on the why and whereabouts of such a claim. The paper was entitled “Sociophysics: a personal testimony.” (Fig. 3.22). At some point, it might be interesting to ask why the founding fathers of a particular field are given the label without having to make the claim for themselves. It could indicate that fathers have to wait quietly until their community recognizes their founding role, which in turn presupposes a wise and fair infallibility of scientific communities in recognizing the breakthroughs within their respective corpus of knowledge. Or, on the contrary it could mean that they obtained the label after an efficient but discrete lobbying within their community. However, there must be many cases of those who are never recognized and who do not make 3.13 Reorientating My Strategy Again to Join a Social Sciences Group 63
Fig. 3.22 After all, why not celebrate my paternity of sociophysics?
the claim themselves so that they just “disappear,” with their contributions maybe staying unknown forever. This is a whole subject in itself that is worthy of further investigation (Fig. 3.23).
Fig. 3.23 It is unfortunate to have to claim one’s paternity butitisevenmore unfortunate to be dethroned from one’s paternity
3.13 Reorientating My Strategy Again to Join a Social Sciences Group
Throughout the whole of the above happenings I never became bitter. I kept up my multiple fights that included the struggle for recognition, the true history and the carrying out of more sociophysics research to deepen my own work. At the same time, sociophysics was expanding in many directions (Fig. 3.24). I then decided once more to change my strategy. The next step was to move physically to join a social sciences group. But that was not an easy task. This 64 3 Sociophysics: The Origins
Fig. 3.24 At the beginning of the twenty-first century sociophysics becomes a well established field of physics. Hundreds of papers are published and several international conferences and workshops are held every year was not easy because the difficulty in principle was not to formally join such a group, although many existed. The aim was to find a social sciences group in which the social scientists were not only not afraid of equations, but rather, they could understand them! In addition, they had to be interested in the sociophysics approach. Surprisingly, such a group existed and so I joined it. The group’s name is “CREA,” i.e., Centre de Recherche en Epist« emologie« Appliquee« and it is part of the Ecole« Polytechnique in France. It is worth underlining that within French institutions, to implement such a move administratively requires the approval of the institutions themselves, which are usually rather strict with regards to the formal rules in place. If such an approval would have certainly been impossible a few years ago, it became quite natural at the beginning of the twenty first century, which is a testimony to the fact that sociophysics has become a well-established field of research in physics. This transformation is not yet directly supported by the institutions themselves but they do not oppose it. On this basis, a move to the social sciences appears as being natural in order to reestablish the formal frontiers of research which is of course an obsolete concept in modern research (Fig. 3.25). My move to my new research world was indeed very successful. I felt very stimulated by the full recognition and the interest I got from the group members. But I must add immediately that this positive support was more local within CREA than at the level of my institution. It did not pay off in terms of my career. But being optimistic in nature, I have not yet given up on the hope that one day it will. When a committee has to evaluate my work, physicists say that they cannot do it since it deals with the social sciences. And clearly it would be impossible for social scientists to evaluate a statistical physics paper. Indeed only sociophysicists could do an evaluation. However, this would require having sociophysics institutionalized 3.13 Reorientating My Strategy Again to Join a Social Sciences Group 65
Fig. 3.25 At the beginning of the twenty-first century, the institution’s hostilities against socio- physics faded away. Some started to be sympathetic to it which will take many more years, a “Catch 22” situation. Interdisciplinary work is always a hard task to be implemented and once it is successful, it is not given the institutional credit it deserves. It is the unfortunate price to pay for bringing innovation to the world of knowledge. But it is worth the price! (Fig. 3.26).
Fig. 3.26 At the beginning of the twenty-first century, the institutions’s hostility against sociophysics faded away
At this stage (in 2011), I am wondering what should be the next move in order to push on with the construction of sociophysics. I am also curious of what will be the next move? Whatever the eventualities, we should keep vigilant on what will happen at the political level since playing politics is also a requirement to reach the institutional level without which the field cannot grow with respect to human power, i.e., to have specific graduate studies, to deliver Ph.D.’s, to obtain tenure positions and so on. The question is then to determine if it is the right time for this institutionalization? 66 3 Sociophysics: The Origins
References
1. K. G. Wilson and J. Kogut, “The renormalization group and the epsilon expansion”, Phys. Rep. 12C, 75 (1974) 2. K. G. Wilson and M. E. Fisher, “Critical Exponents in 3.99 Dimensions”, Phys. Rev. Lett. 28, 240 (1972) 3. B. B. Mandelbrot, “ The Fractal Geometry of Nature”, W.H. Freeman and Company, ISBN 0-7167-1186-9 (1982) 4. S. Galam, “Physicists as a revolutionary catalyst”, Fundamenta Scientae 1, 351 (1980) 5. P. Pfeuty and S. Galam, “Les physiciens et la frustrations des electrons”,« La Recherche JulyAugust, 23 (1981) 6. S. Galam, “Sauver la nouvelle Byzance”, La Recherche Lettre 127, 1320 (1981) 7. S. Galam and P. Pfeuty, “Physicists are frustrated”, Physics Today Letter April, 89 1982) 8. S. Galam, Misere` des physiciens, Pandore 18, 57 (1982) 9. S. Galam, “About imperialism of physics”, Fundamenta Scientiae 3, 125 (1982) 10. S. Galam and P. Pfeuty, “Should God save the queen?”, Physics Today Letter October, 110 (1983) 11. E. Callen and D. Shapiro, “A theory of social imitation”, Phys. Today July, 23 (1974) 12. W. Weidlich, “The statistical description of polarization phenomena in society”, British. J. Math. Stat. Psychol. 24, 251 (1971) 13. S. Galam, Entropie, desordre« et liberte« individuelle, Fundamenta Scientiae 3, 209 (1982) 14. S. Galam, Y. Gefen and Y. Shapir, Sociophysics: A mean behavior model for the process of strike, Journal of Mathematical Sociology 9, 1 (1982) 15. S. Galam, “Sociophysics: a personal testimony”, Physica A 336, 49 (2004) 16. S. Galam, “Pas de certitude scientifique sur le climat,” Le Monde, Mercredi 07 Fvrier,20 (2007) 17. V. Maurus, “Her« esie”,« Le Monde, Samedi 18 Fvrier, Chronique de la mediatrice« (2007) 18. S. Galam, Les scientifiques ont perdu le Nord, Reflexions« sur le rechauffement« climatique, Editions« Plons, Paris (2008) 19. S. Galam, “Entropy”, Semiotext(e) 4, 73 (1984) 20. S. Galam, “Majority rule, hierarchical structures and democratic totalitarianism: a statistical approach”, J. Math. Psychology 30 426 (1986) 21. S. Galam, “Social paradoxes of majority rule voting and renormalization group”, J. Stat. Phys. 61, 943 (1990) 22. S. Galam, “Political paradoxes of majority rule voting and hierarchical systems”, Int. J. General Systems 18, 191 (1991) 23. S. Galam, “Real space renormalization group and social paradoxes in hierarchical organisa- tions”, in: Models of Self-organizationinComplex Systems (Moses), Akademie-Verlag, Berlin, Vol . 64, 53 (1991) 24. S. Galam, Paradoxes de la regle majoritaire dans les systemes hirarchiques, Revue de Bibliologie 38, 62 (1993) 25. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: consensus and attitude changes in groups”, Eur. J. Soc. Psychol. 21, 49 (1991) 26. S. Galam and S. Moscovici, “Compromise versus polarization in group decision making”, in Defense Decision Making, Springer, Berlin, 40 (1991) 27. S. Galam and S. Moscovici, “A theory of collective decision making in hierarchical and non- hierarchical groups”, Russian Psychol. J. 13, 93 (1993) 28. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: II. Conformity and power”, Eur. J. Soc. Psychol. 24, 481(1994) 29. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: III. Conflicts and forms of power”, Eur. J. Soc. Psychol. 25, 217 (1995) 30. S. Galam, “When humans interact like atoms”, in E. White, J.H. Davis (Eds.), Understanding Group Behavior, Vol. I, Lawrence Erlbaum Ass., New Jersey, Chapter 12, 293 (1996) References 67
31. S. Galam, “Fragmentationversus stability inbimodal coalitions”, Physica A 230, 174 (1996) 32. S. Galam, “Rational group decision making: a random 0eld Ising model at TD0”, Physica A 238, 66 (1997) 33. H. E. Stanley, V. Afanasyev, L. A. N. Amaral, S. V. Buldyrev, A. L. Goldberger, S. Havlin, H. Leschhorn, P. Maass, R. N. Mantegna, C.-K. Peng, P. A. Prince, M. A. Salinger, M. H. R. Stanley, and G. M. Viswanathan, “Anomalous Fluctuations in the Dynamics of Complex Systems: From DNA and Physiology to Econophysics,” Physica A 224, 302 (1996) 34. P. Ball, “Utopia theory”, Phys. World October, 7 (2003) 35. D. Stauffer, “Introduction to statistical physics outside physics”, Physica A 336, 1 (2004) 36. T. C. Schelling, J. Math. Sociol. 1, 143 (1971) 37. A. Sokal and J. Bricmont, “Intellectual Impostures”, Profile Books Ltd., London (1999) Chapter 4 Sociophysics: Weaknesses, Achievements, and Challenges
The current trend of physics becoming “global” is analyzed. Underlying the initial hostility of the physics community, I focus on sociophysics in order to enumerate the conditions to establish a new scientific paradigm for the understanding of the human world. Epistemological foundations are suggested so as to be able to provide a framework and to streamline our new emerging field. Existing attempts of Soviet- like rewriting of the history of sociophysics are criticized. Weaknesses and strengths are reviewed with an emphasis on what is intrinsic and specific to the very nature of the field dealing with human beings, from what is structural in the way that the field is developing. Featuring the recent success of sociophysics in the prediction of a few real political events, a strategy is proposed to collectively validate the robustness of the sociophysics approach. The challenge is to provide solid tools to make sociophysics a quantitative and heuristic field of research. A few words of warning: the reader may find several repetitions in this chapter from the previous one. This is done for the sake of completeness of being able to read each chapter independently.
4.1 The Essential Challenges of Sociophysics
During the first years of the 21st century, sociophysics has established itself as a solid field of statistical physics. Hundreds of papers are published in the best international journals of physics and several international conferences and workshops are held every year. Such growth raises three questions about the present status of the field and its future. • Can sociophysics be grounded on some unitary principles and hypotheses? • Can sociophysics identify some major problems to be solved? • Can sociophysics become a predictive field?
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 69 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 4, © Springer Science+Business Media, LLC 2012 70 4 Sociophysics: Weaknesses, Achievements, and Challenges
These challenging questions can only be answered by a respective demonstration of an eventual yes to each question. General analyses could be interesting for their own shake but are irrelevant to the hard science aspect we are addressing here. The future achievements of sociophysics will provide the answers. I have the conviction that the self-organization of people working in the field is a central key to gaining positive answers. But it is far from being self-evident to researchers in sociophysics. This is mainly because of the current trend in physics to be attracted to the “quick and new and jump to the next topic.” We enumerate the conditions necessary for possibly establishing a new scientific paradigm for the understanding of the world of human beings, in focusing particularly on the social and political conditions. Epistemological foundations are proposed to help provide a framework and to streamline our new emerging field. Existing attempts of Soviet-like rewriting of the sociophysics history are criti- cized. Setting a false framework to the emergence of sociophysics will jeopardize the chances of sociophysics in becoming a hard science. Weaknesses and strengths are reviewed with an emphasis on the features which are intrinsic and specific to the very nature of a field dealing with human beings. These are discriminated from those features that are associated to the structural whereabouts of a new emerging field. Accordingly, a strategy is proposed to collectively validate the robustness of the sociophysics approach. This should provide an operative framework for applying predictive tools in order to make sociophysics a quantitative and heuristic field. But the challenge is far from being achieved. I will now make a proposal to confront our involvement in building a solid field of sociophysics.
4.2 Sociophysics: A New Field Is Emerging
Nowadays physics is everywhere. It is becoming “omnipresent.” Or more precisely, physicists are applying their models and tools to an increasing number of different problems taken from almost all fields of knowledge. It has given rise to a series of subfields of physics including sociophysics [1], econophysics [2], networks [3], population dynamics [4], languages [5], evolution [6–8], genetics [9], terrorism [10, 11], and more, [12–16]. Such an interest of physicists toward new fields of investigation that are so distant from the usual fields within physics is a rather new and recent phenomenon. In particular, at the current large scale of activity, there are hundreds of physicists involved, at least part time, in these “exotic” activities. Twenty years ago, most of these applications of physics outside physics did not exist. Indeed, in the early 1980s, a few scarce pioneering papers were available but were looked upon as nonsense by all physicists besides a very few exceptions [17]. Everyone in statistical physics and condensed matter was excited in studying inert matter, enjoying to the full the wonderful world of critical phenomena, and later, disorder and chaos. The physics outside of physics approach remained on the fringe and was even banished up to the mid 1990s. Then, a few clusters of physicists started 4.3 Deciding the Future of Sociophysics 71 to tackle problems from finance using methods from physics under the name of econophysics [2]. Several years later, econophysics slowed down while sociophysics eventually emerged to become one of the main streams of “physics” research of the early 21st century. The change within the physics community has been huge and sudden. The former hostility vanished immediately, with hundreds of papers these days being regularly devoted to these new problems of interest. They are being published in the best international physics journals, including Physical Review E and rarely, Physical Review Letters, some of them devoting whole sections to these new topics, Physica A being a leader. In addition, more than a dozen international conferences and workshops are being held every year on these new subjects of research. Physics institutions worldwide are integrating this “everywhere physics” approach under the internal drive of the many physicists joining the trend. But it also provides a new exciting window to promote physics and counter the decline in recent years in the student attraction toward traditional physics. At the same time, modern scientific societies are in an urgent need of finding solutions, or at least some understanding of a large spectrum of novel social problems produced by globalization and the increasing complexity of the modern world.
4.3 Deciding the Future of Sociophysics
After having revisited the first earlier steps of the current trend of this interdisci- plinary physics, with the initial underlying hostility of the physics community, we address the major issue of its possible future. The focus here is on determining whether or not sociophysics can become a solid predictive tool of knowledge. It is only at an early stage of development but has already reached a critical size in terms of numbers of both papers and physicists involved [5]. There is a motivation to start exploring the possible perspectives and issues which are at stake in the near future. I think that the time has come to start coordinating some of our research activities in order to single out a series of basic questions that we want to address collectively in order to articulate our field and to boost its achievements. Featuring the recent successes in the prediction of a few real political events using models from sociophysics, it is legitimate to consider that sociophysics can help to establish a new scientific paradigm for the understanding of some of the salient features of the human world. To provide an operational basis for fruitful discussion on this crucial challenge, I must enumerate some of the strengths as well as some of the weaknesses of the sociophysics approach. I emphasize and discriminate what is intrinsic and specific to the very human nature of this particular field of research from what is structural in the way the field is developing. I then outline what could be a good strategy for our community for achieving the turning of sociophysics into a new quantitative science of social systems. A proposal is suggested with a few challenges for the future. 72 4 Sociophysics: Weaknesses, Achievements, and Challenges
4.4 Sociophysics: Epistemological Foundations
The phenomenon of emergence of a new field of research is certainly of interest in itself for intrinsic reasons, but the case of sociophysics deserves particular attention for at least three specific reasons. • The first one is the very unusual and unnatural connection in sociophysics between the two fields of research that are the furthest apart: that of the most complex field concerning both social structures and dynamics of human beings, and the most primitive one, that of inert matter and atoms. • The second reason is the potential and huge danger which could result from a misuse of sociophysics in the political world. • The third reason is the potential and huge benefits, which could result from an appropriate use of sociophysics in the political world. Dealing with human behavior in the name of science may become dangerous in both the opposing cases of a valid model and a wrong model. Even if a model is totally wrong, it can always be used as a false argument to support a biased political view. At this level of misuse, the intrinsic authority of a supposedly scientific proof can turn out to be devastating since it would be impossible to be refuted by most people. This is not the case of a regular social science-like argument, which by its very nature, can always be either refuted or at least rejected as not being convincing. At the same time, if a model is eventually proven to be valid, it may provide an efficient tool for manipulating people. Therefore, in both cases of wrong and valid models, some caution is required to clearly define the content and the limits of sociophysics. It is the prerequisite condition to have it developed in a coherent and scientific frame. It is worth stressing that, contrary to sociophysics, applications of physics to biology and finance present no obvious danger. Either some model is valid, even in part, and then can be useful to some extent in its applications, or it is wrong and at most, results in some loss of money for the people who were expecting to benefit from it. So there is nothing much to be worried about. However, this is not the case in sociophysics since it addresses problems directly connected to human beings of which the consequences could be dramatic. On this basis and since the very name of sociophysics induces certain cognitive representations of what it could be, it is an essential and instrumental priority to first define what sociophysics is and even more importantly, what it is not. At least, this is my conviction. This does not imply that every sociophysicist should agree with me, but I seek to provide the first attempt to build some epistemological foundations for our new emerging field. Afterward it will be up to the community of sociophysicists to construct a stimulating and safe framework by setting the limits and the goals of the discipline. These are the prerequisites for the establishment of a solid science. I, therefore, propose ten guidelines in which five are positive and five are negative. The positive statements are about what sociophysics should be. 4.4 Sociophysics: Epistemological Foundations 73
(1) Sociophysics is the appropriation by physicists as physicists, of the study of some parts of social and political behavior. It is a novel additional framework within which to tackle problems related to human behavior. Some are identical or similar to the problems studied in the social sciences but some are new. The point of view and interest with respect to each field are different and specific. Connections and overlaps with the classical social sciences may appear and are welcome but they are neither the rule nor the ultimate goal. Sociophysics is merely the task of physicists. The physicists working in sociophysics do not have the vocation to become social scientists in the usual sense. They aim at becoming “sociophysicists.” Social scientists are requested to learn a little bit of statistical physics. (2) The novelty of sociophysics lies in the process of modeling social behavior in the same way that physics models natural phenomena. It is not bound to proving the universal nature of a solution, but focuses on finding out the simplest conditions which can reproduce a universal feature of a given phenomenon. This is the fundamental and irreducible difference from a purely mathematical approach. (3) The purpose of sociophysics is to discover particular trends in human behavior while ignoring many other aspects of it. The goal is to grasp a quantitative understanding without accounting for the full description, which would involve all the details. Sociophysics should not be taken literally at the level of its numbers although its models yield precise numbers. Trends and dynamics are the important features to rely on. (4) Sociophysics is a mosaic of models, which can be either complementary or even competing and contradictory, each class of models tackling only a very small part of reality. It is an evolutionary and partial description of the social world, but is not itself the reality. It is at its very earliest stages of development. (5) It aims to discover some empirical laws which can be tested against real data to describe social and political behavior from another viewpoint, which is not exclusive. In addition to the above five positive statements, the following five negative statements can be enumerated in order to set the landmarks for sociophysics to be able to develop. The negative statements are about what sociophysics should not be. (1) Sociophysics does not aim to establish absolute truths, which would govern all aspects of human behavior. It does not pretend, either, to have a religious-like status nor to become the ultimate theory of society. (2) Sociophysics does not focus on the search for some mathematical paradigm to the description of the social sciences. It does not emphasis the use of mathematics in describing human behavior. (3) Sociophysics does not pretend to be the first attempt ever to quantify social behavior using either mathematics or statistics. (4) Sociophysics does not aim at establishing theorems in the way that social scientists have been using mathematics for many years, as for instance in the extensive use of game theory in both the economic and political sciences. 74 4 Sociophysics: Weaknesses, Achievements, and Challenges
(5) Sociophysics does not intend to substitute itself to the classical approaches used in the social sciences. It does not replace the traditional studies by social scientists and does not claim any superiority over them. The above ten statements are proposed for establishing sociophysics as a field of research in itself. It is clearly open to discussion and can be amended. Its achievements will certainly contribute to shape it more adequately. It is the first explicit framework that aims to set up an epistemological foundation to our discipline. It has the vocation to be adjusted to the outcome of its own research.
4.5 Flashback to the Origins
In recent years, the physics-like modeling of social and political phenomena has been the subject of a growing number of papers from statistical mechanics and condensed matter physicists. The trend has become significant in terms of numbers of papers and international conferences and workshops. Although the field of Sociophysics is still at an early stage of development, it is creating a great deal of interest and could soon reach a critical size, at which point it could start either to decline or to turn into a new science. The perspectives and stakes are as numerous as the dangers and possible misuses. At this stage, to get a better understanding of what sociophysics is, and to optimize the chances of it succeeding in becoming a solid quantitative science, it is useful to trace back its origins within physics. I have already given [17] a personal testimony on the earlier stages of socio- physics, which was initiated in the late 1970s and beginning of the 1980s. It is worth recalling the historical context of condensed matter physics at that time. In 1971, after decades of theoretical difficulties with the ongoing failure to explain the fascinating puzzle of phase transitions, all of a sudden, Wilson, a nonprolific physicist in terms of number of publications, suddenly published two papers [18] with the so long awaited solution. One year later, Wilson and Fisher made the theory practical for calculations with the so-called -expansion [19]. The solving of the enigma of critical phenomena opened what can be described as a “golden age” of physics in terms of quality and quantity. The concepts and tools of the renormalization group techniques opened up a wide area of intense and active research, which has involved hundreds of physicists and the publication of thousands of both theoretical and experimental papers. The concepts of universality classes, irrelevant variables, and critical dimensions, lower and upper, were of a fantastic excitement in creating an incredible momentum in physics research. At that time, in the mid-seventies, my advocating of sociophysics as a natural extension of condensed matter research among physicists was considered as being pure nonsense by everyone except a very few exceptions. Moreover, such an option was perceived as a threat that could jeopardize the tremendous status and strength of physics at that time. Claiming that statistical physics could be applied to economy, 4.6 The Soviet-Like Rewriting of the History of Sociophysics 75
finance, sociology, politics, and psychology prompted a unanimous hostility from physicists. Such a reaction was driven by the elitist contempt physicists shared toward the social sciences in general. During the following years I kept on trying to spread my convictions amongst physicists, without much success. In fact, what I was doing was envisioning the tremendous perspectives of such a crossing of fields for the understanding of human society. It was obvious to me that on the one hand we had some new powerful tools and concepts, and on the other, there existed a great poverty in the understanding of social behavior. I was basing my call for the creation of this new field as a natural development of physics on both its internal limits as a field of research, and on the growing contradictions of physicists as a social community. Being too skilled to only tackle complex technical problems, they will ineluctably be tempted to play around with their so powerful “toys” in other areas. On this basis, I developed a coherent analysis with a dynamic perspective of a sociocultural phenomenon [20–22]. I foresaw the emergence of econophysics, sociophysics, and other “X” physics as an inevitable and deterministic outcome of the dialectical transformation of the internal contradictions, which would arise from the interplay between the subjects of physics, the tools of physics, and the people doing physics. I viewed it as a natural and direct outbreak of the last major and fundamental success of physics as a whole field of research. But right ideas can stay confined to a few people for a long time before they eventually propagate. In my case, they became plausible and attractive for others only in recent years.
4.6 The Soviet-Like Rewriting of the History of Sociophysics
With sociophysics being an active and recognized field of research, but not yet a mature one, only a few books [12, 15, 16] and review papers [1, 5] have been published so far. But it is significant to notice that the ones in which the history and development of sociophysics are discussed are adopting the same politically correct but false viewpoint. They are articulated around three central preconceptions, which are: (1) The refusal to admit its conflicting aspects within the physics community. (2) The discarding of the first founding papers, that were controversial by their very existence. (3) The artificial link to the social sciences with the construction of an imaginary perfect and smooth path of its emergence. Accordingly, sociophysics is given a biased positioning so as to fit a nice, linear historical chronology from some fake origin, while removing all the controversial aspects. Because sociophysics is now accepted by research institutions and by most 76 4 Sociophysics: Weaknesses, Achievements, and Challenges physicists, all of the “negative elements” are deliberately ignored. Only the harmony of the research community is underlined. In this way, it is necessary to leave to one side the perturbing physicist I have been at the earlier stages of sociophysics. I will focus on two publications which are emblematic of this Soviet-like rewriting of the history of sociophysics. One is a paper by Stauffer, “Introduction to Statistical Physics outside Physics” [23] and the other is a book by Ball, “Utopia theory” [24]. While the two address the question from a different perspective, discussing rather different aspects, they both adopt the same politically correct posture to legitimate the presentation of sociophysics by some falsely reconstructed history linking this field to some past work in the social sciences. They even argue about its relevance, acceptance, and rejection, but at the same time they totally ignore my own many founding contributions. I am not mentioning these facts to argue or to express what would be a legitimate bitterness. I am emphasizing it to illustrate how in general, writing about science history can be misleading and even wrong. What we have here is a very good example of what can happen with respect to other questions of science breaking into other fields. It is particularly because I do not take this as solely a personal matter that I comment on it at length. To build a history of the breakout of physics outside of physics by physicists, they present both a selection of isolated historical facts, which in turn are used to trace back some linear path to reproduce their own view of a happy and smooth view of the dynamics of science. Simultaneously, they deliberately ignore certain facts, seminal contributions, and instrumental events. Not only do they totally ignore the very conflicting nature of sociophysics with respect to the traditional social sciences but they also silence the initial stiff opposition of the physics community to it. Sociophysics is thus presented as a natural and smooth outgrowth from the social sciences. To ground their a priori view, and to manipulate the real history, they both trace back the phenomenon of sociophysics within the social sciences by setting an organic link to the work by Schelling in 1971 [25]. They claim he was already applying physics, although without knowing it. Along this line of thinking, a few papers have tried to show that the so-called Schelling model was somehow the Ising model [26]. But it is different from the Ising model. The Schelling result for urban segregation is based on a rudimentary simulation done-by-hand using a mixture of two species, at fixed densities. It is a multiagent-like simulation, nothing less nothing more. That does not make its finding uninteresting but neither make it a foundation of sociophysics. Although no explicit link to physics was mentioned, Schelling is given a kind of “innocent paternity” to sociophysics. Moreover, these pseudohistorians do not comment on the fact that Schelling’s work has been anecdotal in being totally ignored by physicists, including myself, up until recently. Nevertheless, it is an interesting contribution, which validates the compatibility of sociophysics with the social sciences. The above “sleight of hand” trick is remarkable in apparently legitimizing the sociophysics approach as being a natural extension from the social sciences itself. 4.7 Fatherhood with a Touch of Humor 77
By doing so, all the conflicting and controversial parts of it, with respect to the social sciences and within physics are simply erased. It is thus unsubstantiated from its necessarily conflicting and controversial features. It is important to underline that sociophysics has been developed by physicists within physics and independently of influence or link to earlier social sciences work. It is essentially an internal development among physicists, which does not preclude a qualitative change in the future. Up to now, sociophysics has grown essentially within physics. This fact is both its strength and its weakness. Another typical attempt of such a revisionist construction of the genesis of sociophysics can be found in a paper by Stauffer and Solomon [27]. Of course, pushing along this fictional kind of approach, one could state that in fact it was even Aristotle who initiated this topic of sociophysics a good few centuries ago. From the perspective of today, he could be regarded as a physicist and since he also lectured about politics, he must have been in some way a sociophysicist as well [28]. Soon we will be using physics to study atoms, which will appear as being a breakthrough from “antique” sociophysics! Besides my rather bruised ego, it is important to state the real facts even if they are crude and disturbing. This should matter especially for historians of science. I am stepping in to denounce this false reconstruction of history for mainly three reasons, which are: • It is of importance to state the real but crude facts, which should particularly matter to scientists, while not giving up the human content of the history. This is all the more so, since this human aspect reveals some fundamental questions about the very nature of sociophysics and its development. • It is crucial to oppose an epistemological falsification of sociophysics while it is still possible, since most of the few protagonists are still alive. • Claiming a wrong idealized origin to sociophysics may well lead to a total misconduct in the future applications of physics outside physics.
4.7 Fatherhood with a Touch of Humor
Bearing in mind the above discussion, I do claim to be the father of sociophysics, without any false humility. Nevertheless, I immediately add that I am fully aware and I acknowledge that: • As is well known in human reproduction, to be the father does not mean to be either the first one nor the unique one to have had intercourse with the mother. • Moreover, as with any intellectual matter, no one is ever the first one to start from scratch. As wisely stated by the former King Solomon “Nothing is new under the sun”. • The question of the mother, if any, is open! • Last but not least, the existence of a father presupposed the existence of a grand- father whose identity could be multiple. And so on and so forth. 78 4 Sociophysics: Weaknesses, Achievements, and Challenges
Accordingly I am not claiming to be the first one, who has thought or done something, which can be connected to sociophysics. I clearly do not ignore previous work associated to the field. I have even cited this work, that which I knew of, in the 1982 manifesto of sociophysics entitled “Sociophysics: A mean behavior model for the process of strike” [29]. The paper contains a whole argumentation about why and how the crossover to other fields should be justified. A basic discussion about the ethical implications of such research is also included. This founding paper, which I cite, was co-authored with two other young physicists, Yuval Gefen and Yonathan Shapir. This fact does not prevent me from standing as the founder of sociophysics. The reason for this is that it is not a single contribution that matters, but a series of contributions together with an explicit vision clearly framing the field. Among the earlier contributions cited in [29] is the 1971 Weidlich paper “The statistical description of polarization phenomena in society”. He used Master equations in the spirit of Lotka–Volterra-like population dynamics to study opinion dynamics [30]. Weidlich named his approach sociodynamics and stated that it was part of synergetics, a field developed by Haken [31], which has not much to do with sociophysics. Later on, Weidlich kept on developing his work, continuing to use the Master Equation approach [13]. Accordingly, it cannot be defined as a founding contribution to sociophysics. I would like to take the opportunity in this book to cite an unknown short- note by Stenflo and Wilhelmsson published in 1981 entitled “Nonlinearities and soliton-like structure in society” [32]. I was ignorant of its existence for many years. Later, Stenflo told me that at the time, it was the condemnation of some of their colleagues that prompted them to give up pursuing the investigation of their idea of a “plasmasociology.” It made their note a one shot, isolated contribution. I am the father of sociophysics not because of one or two earlier papers, or because I would have been the first one ever to think about applying physics to social behavior. I founded sociophysics as a field simultaneously from a philosophical viewpoint, an epistemological analysis, a consideration of physicists as a social body, and with the concrete contributions of building a series of models, which use some basic statistical physics models to deal with a large spectrum of social and political problems. In addition, I have been an active militant advocating the cause for 30 years among my colleagues, as well as those outside physics. During these earlier pioneering years, I thus organized a series of seminars along this line to try and get others to join in my vision. • The first series was at Tel Aviv University with Ben Jacob. We setup a “Bullshit Seminar” at the physics department from 1978–1980 to address all these questions from a “revolutionary perspective” within the physicist’s community. The issue at that time was to blow up the rigid frontiers of physics to extend our field of investigation from nonliving matter to the living worlds. Ben Jacob was advocating the use of physics in biology, and myself, its use in the human world. The seminars generated a lot of interest, as well as polemic and even attacks. It was a success in terms of public and impact. 4.8 Basic Weaknesses of Growing Sociophysics 79
• I held the second series of seminars at NYU during the years of 1981–1983, gathering together physicists, economists, political scientists, computer scien- tists, and psychologists. Besides having an interdisciplinary talk, we used to chat around a cheese and wine table. It yielded a great deal of interest and exchange. Maybe the good wines were to blame for its success? • I repeated the experience again in Paris during 1991 at the Pierre and Marie Curie University with most of the few people interested in the approach in France. Surprisingly, there was no cheese and no wine, only passionate discussion. So maybe the wine was, after all, “innocent” in the seminar success, although rather enjoyable, at NYU. It is thus my feeling of unfairness and injustice of not being given the legitimate credit for my vision, my fight, and my contributions to the birth and establishment of sociophysics, which drives me to take such an uncomfortable posture in claiming its paternity. The task may even appear ridiculous, given that the field is growing, with many people not citing me adequately for my series of anterior founding papers. It even made me totally paranoiac when I found papers appearing in leading journals such as the Physical Review Letters based on my exact models that were presented as being novel. For one of them, I had to exchange several emails to have the authors recognize the fact and have this corrected in a follow up paper in the Physical Review E. But who will connect the two papers within the huge inflation of the number of published papers? What is seen as mattering is the first publication in the prestigious Physical Review Letters. This is the one which is cited, giving automatically the credit to its authors. However, they are entitled to claim that they have been cleared from misconduct since they acknowledged the omission later on. It is a very efficient “technique.” which cannot be found out by just reading the papers. Somehow everyone is beyond reproach, those who missed crediting the original work since they did acknowledge the fact later in a following publication, as well as those who cite the publication that claimed to be first since it is the most visible one. The unfair benefit will thus prosper forever. I could cite names and references but what for? I do not intend to make the issue personal, although I am personally bruised and I do not aim either to hurt anyone. I am also aware that reporting on these practices may make me appear pathetic and bitter, but yet, like in the song: And I said to myself what a “wonderful” world! I decided not to keep silent even at that price, because it is of importance to make visible the hidden part of the “wonderful” world of research. I am convinced that such practices happen very often in all fields of fundamental research where there are no patents that would allow possible legal investigations.
4.8 Basic Weaknesses of Growing Sociophysics
Restoring the truth about the past history is just as important as scrutinizing the current situation of sociophysics in order to forecast possible evolutions, which are nevertheless only conjectures. 80 4 Sociophysics: Weaknesses, Achievements, and Challenges
At the beginning of the 21st century, sociophysics is a flourishing field of statistical physics. An increasing number of physicists are joining the field and all physics journals accept related papers. Every year several conferences include sociophysics topics. Summer schools and workshops contribute a good deal to the growth of this new field of research. Outside physics, several social scientists are starting to become interested, although with a certain amount of doubt and caution. Journalists are usually interested, and several articles are published regularly on associated subjects in major newspapers around the world. On this basis, the strategic question is to determine if sociophysics can really become a predictive social tool. However to figure out even the beginning of an answer it is necessary to analyze what are the respective strengths and weaknesses within the implementation of sociophysics in terms of research papers and solid achievements with respect to the real world of social behavior. Among the major weaknesses stands, as could be expected, what is currently happening in all fields of research. Namely, there is a dilute clustering of publica- tions in a kind of fragmentation within informal lobbies, each ignoring what the other is doing. No credit is usually given to the relevant papers, and earlier work is republished as being new. Plagiarism, conscious, or unconscious, is rife. However, this aspect of research is happening in the emerging new field of sociophysics at a wider scale. At the time of the internet, if a paper is not readily available, then it does not exist. Moreover, if it was published too long ago, then it exists even less. The common argument to justify such a practice is the impossibility of keeping track of the millions of papers being published these days at very high speed. What is not delivered within the first two pages of a Google search does not exist. In addition, no sense of guilt or wrongness is considered since each author feels that they do not receive the credit that they deserve from others, and, therefore, cannot be blamed for not giving the necessary credit to the relevant papers they did not know about. Moreover, the imperative of searching and finding out about anterior work no longer seems to be a requirement for scientific activity. It is a serious change in research practice in which only very short-term and restricted memory is becoming the rule. In the long run, it can only jeopardize the solidity of scientific research. But to analyze such a new trend further is out the scope of the present book. It is unfortunately a general feature of modern electronic publishing of scientific research. Tackling this fundamental change in the way science progresses will be the major challenge of this new century. Within our field of sociophysics, other specific misconduct exists. For instance, in a given paper, once a paper is referred to from a social sciences journal, it is put forward to produce a social sciences legitimacy to the corresponding work, even if it is not connected to the work itself. Then, everyone recopies the reference without reading it. It becomes the “natural” social source of the problem treated by the physicists. Such practices will not help in making sociophysics a strong and robust field of science. Up until now, several models have been elaborated together with new concepts and many numerical simulations. While they are claimed to describe a series of 4.9 The Positive Achievements of Sociophysics so Far 81 social and political situations, most of the work, including some of mine, is indeed totally cut off from any real system. Moreover, each one focuses only on its own idea, often limited to a local update rule motivated by a qualitative analogy with a rather general social feature. In addition, the same models are often rephrased and republished as new ones. It has happened several times with my own models. But although I may appear paranoiac, as mentioned earlier, I do not believe that everyone is against me. So I do not take it personally, but on the contrary I look at it as a sign of a serious failure of our growing field of research. All these ill practices are the direct result of the fact that at the moment no unified corpus is being constructed for sociophysics, and not even a theoretical framework. There exists much research in the field, but mostly carried out by separate groups or individuals, who often totally ignore previous research. No agreed criterion is available to validate a model. No single specific social problem is clearly identified as “the problem” to be solved in the field. There are only claims of relevance to hypothetical applications that are made here and there. As mentioned above, the most “hip” thing to do is to cite some old papers by a social scientist to create the illusion of an organic legitimacy of the approach to the traditional social sciences and at the same time to show off knowledge of historical culture.
4.9 The Positive Achievements of Sociophysics so Far
First, it is worth underlining that up until now, sociophysics has not led to any negative achievements. We can then question whether there exists any positive ones? Indeed, what has sociophysics accomplished so far, if anything? The answer is yes, sociophysics can already boast of a series of achievements, although not all on the same footing. Here is a nonexhaustive list. • Several models have been elaborated together with new concepts and many numerical simulations to shed new light on several social phenomena. But these results are at a rather qualitative and formal level. • Some general qualitative features and properties of both opinion dynamics and voting have been given new points of views. However, as for the preceding item, these viewpoints are still very general. • Some past political events have been revisited with unexpected explanations. – A voting model was developed that provided a key to understanding the collapse of the communist party in the last century [33–38]. In particular, it yielded a coherent explanation to their sudden collapse following decades of frozen leaderships. Indeed, the model shows how a very long and sustained increasing opposition within the party can have no effect at all for many years before all of a sudden it turns the leadership up-side down without any warning. 82 4 Sociophysics: Weaknesses, Achievements, and Challenges
– A contrarian model was proposed to explain the fifty–fifty 2000 American and 2002 German elections after the events [39–41]. Using this model, it was then advocated that if it is valid, then more fifty–fifty outcomes should occur in the near future in democracies. – The so called Sznajd model [42] was used to explain the 1998 Brazilian elections [43]. – A social percolation model proposes the basics of a global framework con- cerning terrorism events such as the September 11th attacks in 2001 [10, 11]. – A spin glass-like model was used to explain the European stability during the cold war period as well as the Eastern European instabilities, which followed the dissolution of the Warsaw pact. This model applied in particular to the ex- Yugoslavia. The recent increase of EU members was also suggested [44, 45]. – The Indian elections have been investigated [46]. – The distribution of the number of votes received by candidates in proportional elections was shown to obey some universal scaling function, which was identical in different countries and years [47]. However, while coming up with an explanation to a past event is satisfying, it is not really convincing, and more so when a different model is used for each event. One step further has also been accomplished; that of predicting certain events that occur in the near future but without mentioning a date. • Using my voting model, I predicted how the extreme right party in France, the National Front, could eventually arrive to power democratically and by total surprise. No one believed it, but what was predicted did actually nearly happen. The National Front scenario did occur in part in 2002 with its leader running in the second round for president, to the total surprise of everyone, including the NF itself [48, 49]. • Using my contrarian model, I predicted that fifty–fifty elections were about to occur again and become a common feature of western democracies. No one believed fifty–fifty elections could happen again. But what was predicted did happen. Fifty–fifty elections did occur again in the 2005 German, 2006 Italian, 2006 Mexican, and Czech elections in contradiction to all the polls and analysts’ predictions [39–41]. • Elections in Germany and Bavaria were investigated by Schneider and Hirtreiter who not only explained the past, but also predicted that the CSU will rule Bavaria forever, if the current political structure remains valid [50, 51]. No one gave a serious thought to the above successful predictions, being too busy trying to recover from the psychological and political collapse which resulted from these totally unexpected political events. And, even if it sounds nice, they are still not completely convincing arguments to validate the approach since in both cases neither a date nor a precise location were given. It is always easy to say “I told you so!” after the event. Before ending this part, let me point out a worthwhile digression that high- lights the existence of the remaining restrictions from physics institutions toward 4.10 The First Sociophysics Successful Prediction of a Precise Event 83 sociophysics. My contrarian paper was first submitted to PRL where it was withheld by the editor, who refused to send it to referees, arguing that it was too political. Later on, the Physica A Advisory Editor Marcel Ausloos showed a much more open mind and accepted the processing of the paper. It was refereed and was eventually accepted after some revisions.
4.10 The First Sociophysics Successful Prediction of a Precise Event
Along the same lines of a general prediction, using my minority spreading model, I made the specific recommendation about the temptation to hold a referendum in the process of the European construction within European countries. In my 2002 paper, I concluded with the following words [52]: To give some real life illustrations of our model, we can cite events related to the European Union which all came as a surprise. From the beginning of its construction there has never been a large public debate in most of the involved countries. The whole process came through government decisions though most people always have seemed to agree on this construction. At the same time European opponents have been systematically urging for public debates. Such a demand sounds absurd, knowing that a majority of people are in favor of the European Union. But anyhow, most European governments have been reluctant to hold a referendum on the issue. At odds with this position, several years ago, the French president Mitterand decided to run a referendum concerning the acceptance or not of the Maastricht agreement. While a large success of the “Yes” vote was taken for granted it indeed made it just a bit beyond the required fifty percent. The more people discussed the issue, the less support there was for the proposal. It is even possible to conjecture that an additional two weeks extension of the public debate would have made the “No” vote win. I then reiterate the conclusion in 2004 stating in the conclusion of another paper that [53]: Applying our results to the European Union leads to the conclusion that it would be rather misleading to initiate large public debates in most of the countries involved. Indeed, even starting from a huge initial majority of people in favor of the European Union, an open and free debate would lead to the creation of a huge majority hostile to the European Union. This provides a strong ground to legitimize the on-going reluctance of most European governments to hold referendums on associated issues. It is of importance to underline that when these statements were made, no referendum was planned in Europe with respect to the proposal of a European constitution. 84 4 Sociophysics: Weaknesses, Achievements, and Challenges
4.10.1 Taking Risks to Validate Sociophysics
Then, at the end of 2004 in France, Jacques Chirac decided to hold a referendum about adopting the project of the European constitution. This turned out to be a great opportunity to make a well defined and precise prediction. Indeed, if it is nice to produce explanations of past opinion formation issues, and sociophysics has proven to be able to do this, it would be much more convincing to predict an outcome of a future opinion issue. This referendum was the perfect opportunity to apply my opinion model to make a prediction concerning the outcome of the referendum. The polls were giving 20% to the “No” vote, 70% to the “Yes” vote and 10% to abstentions. It was taken for granted by everyone that the “Yes” vote would win the day, including those from the “No” vote. The unique issue was the degree of participation in the vote, many abstentions being feared, which could weaken the expression of the French people’s support for Europe. In the mean time, I was able to introduce the existence of heterogeneous beliefs to make the minority opinion spreading model applicable to more fuzzy issues with different subpopulations [54]. I then made the analysis using rudimentary investigations by myself. My conclusion was that the critical threshold for the “No” vote to start to inflate as a result of public debate was located in the vicinity of 15%. I also concluded that a long time would be necessary for the “No” vote to pass over the 50% mark. Several months ahead of the vote, before the official campaign started, the “No” vote was scoring around 20% in the polls. There were five months of debate ahead. Therefore, from the model, within the given conditions of the debate during this period, the prediction was that the “No” vote would eventually win the day. This outcome was quite clearcut; the only problem was that I could not believe it myself. A huge majority of people were in favor of the “Yes” vote, almost all political leaders were in favor of the “Yes” vote, all the media were in favor of the “Yes” vote, all the stars were in favor of the “Yes” vote and still other vociferous groups were in favor of the “Yes” vote. For almost everyone, France could not say “No” to Europe; that would have been totally absurd. And I agreed with this. It happened that I was interviewed about my minority spreading model and its application to the coming referendum by a journalist from the newspaper Le Monde [55]. At the end of the interview, he asked me “Are you sure you want to have your conclusion printed in black and white? Your theory is nice, but the conclusion is nonsense and you will lose all credibility in the future”. For a minute, I became really scared, asking myself “Why announce an event that I myself do not believe in? Why exhibit myself in front of millions of people as a fool?” As the journalist told me “Printed matter lasts for a very long time and will be used whenever necessary against your approach. Think about it.” And I did. I realized that what this was all about was at the very core of the sociophysics challenge; not to try to win a personal reputation, but to build a robust theory of social behavior with a heuristic power, based on a scientific procedure, which can be tested against real-life events. It was not and should not be a personal issue. 4.10 The First Sociophysics Successful Prediction of a Precise Event 85
A solid social science could be built only by making predictions, which could be tested against reality. If a prediction turned out to be right, it would validate, at least in part, the model. If the prediction turned out to be wrong, the reasons for the failure would provide either more insights into the model, or lead to an abandoning of part or all of the model. This is how we proceed in physics. So, the article in Le Monde was published with my very clear prediction of a “No” victory [55].
4.10.2 When the Prediction Turns Out to be True
On May 29, 2005, the “No” won the vote, with a score of 55%. It was a total blow to all the French elite. Up until the last minute, no-one among the analysts could believe that it would happen. But it did. This is an important case that goes to show that it is not always the media that makes public opinion. I myself had contradictory feelings as a result, with on the one hand, the legitimate satisfaction of having forecast such an event using my opinion dynamics model, and on the other, the fear of how the European construction would find a way out of such a failure. The new simplified treaty which was set up later resolved what could have been quite a setback for Europe. Coming back to the meaning of my successful prediction of the “No” victory, it is worth emphasizing the following points: • This particular case study was the very first time that the outcome of a political vote had been successfully predicted using a model from sociophysics. • Moreover, it was a highly improbable event. It was not a heads or tails chance. It was not even a random selection with a low probability. Up until a few weeks before the vote, it was a zero probability event. • In addition, the prediction was made several months before the actual vote, in contradiction to all the polls and other predictions. • On top of this, I predicted a few years earlier that the use of a referendum to strengthen the European construction would be counterproductive, leading to a reversal of the current large public support. It is also of importance to add some points of safeguard: • A single successful prediction is not enough to conclude that the model is correct. More tests are required. • However, the peculiarity of sociophysics dealing with human behavior makes it impossible to reproduce exactly the same “experiment” or real social event. Therefore, it should be a series of successful predictions for similar events such as different referendums, which must be the criterion for solid validation. Nevertheless, this unexpected successful prediction at least validates the model in the sense that it is worth continuing investigating it. It also justifies the hope that sociophysics in the future may indeed yield real predictive tools. 86 4 Sociophysics: Weaknesses, Achievements, and Challenges
At the same time, it is of central importance to clearly note that, if it is proven to be true, i.e., sociophysics becomes a solid predictive science, it will have drastic and unknown consequences on our social and political lives. No one is currently prepared for such a possibility.
4.10.3 When a Prediction Fails
Motivated by the success of my 2004 general prediction of the increasing occurrence of fifty–fifty elections in world democracies, with several cases including Germany, Italy, Mexico, and the Czech Republic, in September 2006 I had an article published in the paper, Le Monde [56] about the possible outcome of the May 2007 French presidential election. I warned against the danger to the French social stability of a fifty–fifty election in the case of a Royal–Sarkozy second round. Nevertheless this prediction was not as precise as for the 2005 referendum. I mentioned it as a possibility since at that time the list of the running candidates was still unknown and not official. Moreover, even the programs of the respective candidates were unavailable. However, the second run did occur with a Royal–Sarkozy standoff and Sarkozy did win with a substantial margin. No fifty–fifty scenario occurred. This failure prompted me to react in the way that scientists do, examining the basic hypotheses of the model. This study indicated firstly that contrarian behavior is not systematic, but is only a function of the general state of society. Different situations may produce different levels of activation of contrarian behavior. Secondly, this misfit sheds light on the fact that other types of behavior should be included in our model of opinion formation. Accordingly, we have investigated the effect of having inflexible agents to discover some similarities and differences with contrarians [57].
4.11 Proposal to Establish a Road Map
The successes and failures mentioned above clearly legitimate the sociophysics approach and simultaneously indicate that much more research has to be carried out. The paradoxical aspect of the growth of sociophysics is that physicists doing sociophysics would like to have social scientists become interested in their work. They complain that social scientists do not read their contributions published in physics journals while they try to establish a somewhat fictional organic link to the social sciences by citing one or two “legitimating” papers. Then, they have probably never read most of these papers but only recopied the references from one paper to another. At the same time, they do not read my earlier inaugural papers because either they have been published in nonphysics journals, which are not immediately at hand, or they were published too many years ago, outside of their memory that extends only to the last few months. 4.12 What the Climatologists Did with the IPCC Should Not Be Repeated 87
Moreover, while hoping and calling for a real link with social scientists, they have mostly ignored my extended work of collaboration with the social scientist, Serge Moscovici, which is a rather rare case, that deserves more attention. All these facts are given to show up some of the aspects of the real dynamics of scientific research, which are rarely given so explicitly. It underlines the current trend of the scientific community to produce papers by following the rather transient fashion of moving at quite a quick pace. With respect to sociophysics, I prefer to be optimistic and consider that all the mentioned weaknesses are the direct outcome of its childhood. The debate is open, as is the future of sociophysics. I think it is time to reduce isolated work, or more precisely, in addition to whatever work is being carried out to focus on a precise list of real events in order to try and predict them, so as to be more efficient and fruitful. I make the following proposal to the sociophysics community: I propose selecting collectively some real political issues that will take place in the next three to five years, and then everyone interested applying their respective models in order to make predictions about them. Once the results are known, each model could be evaluated according to its respective overlap with the outcomes of these real events. Both what was right and what was wrong about the outcomes and the models could then be combined to determine the next step in our research.
4.12 What the Climatologists Did with the IPCC Should Not Be Repeated
However, if it is of importance to start to coordinate the various projects carried out in sociophysics, it does not mean that we should set up an organization to monitor it. It is of the utmost importance to stress that any temptation to reproduce the “successful” precedent of climatology would certainly open a Pandora’s box towards a new kind of totalitarianism. Why? Twenty years ago, climatology was far from being in the headlines of research. Neither much money nor many jobs were allocated to what was then a minor discipline. At some point, climatologists set up several national and international structures to both coordinate and to unify their activities. Indeed they managed to come up with a very efficient network which turned their discipline into a priority subject at the beginning of the 21st century. They were able to embody international institutions including the United Nations as well as major national agencies all over the world. Taking advantage of their dominant position, they went on to predict certain catastrophes, together with the corresponding policies that needed to be takeninordertoavoidthem. However, while the initial motivation was praiseworthy, the net result has become a politically oriented machinery which makes false statements in order to reach some ideological goals. The whole framework has been wrongly claimed in the name of science and scientific proof [58–61]. 88 4 Sociophysics: Weaknesses, Achievements, and Challenges
To imitate such a construction would be disastrous both for sociophysics and humanity. At first it would give definitive strength and means to sociophysics but then it would unfortunately lead to the determining of a single voice by some institutional body in order to settle the “scientific” guidelines for a perfect society. In a similar way that today, climatologists update their advice to policy makers, sociophysicists would dictate their demands to politicians. And, as has happened with the opponents of climatology, the opponents would be denied the right to contest the official truth established by the representative institution. However, as the context would be political, the chances are that rather quickly, a dictatorship would emerge, founded in the name of science. Therefore, any sociophysicist should adamantly oppose any attempt to establish a single body to deliver a unique voice for the field. Therefore, while scientists must be engaged in the various public debates, and should popularize their research and their suggestions, they must do it individually and never as an organized group. Such a practice will guarantee the free and contradictory expressions of different scientists. It will avoid the politicizing of science in which some scientists would be tempted to confiscate certain aspects of science in order to promote their ideological view and preserve their status and power. Otherwise, the social consequences could be really dramatic.
References
1. S. Galam, “Sociophysics: A review of Galam models”, Int. J. Mod. Phys. C 19 409–440 (2008) 2. R. N. Mantegna and H. E. Stanley, “An Introduction to Econophysics”, Cambridge University Press, England (2000) 3. A. L. Barabasi´ and R. Albert, “Statistical mechanics of complex networks”, Rev. Mod. Phys., 74, 47 (2002) 4. M. Droz and A. Pekalski, “Population dynamics with or without evolution: a physicist’s approach”, Physica A 336, 84 (2004) 5. C. Castellano, S. Fortunato, V. Loreto, “Statistical physics of social dynamics”, Rev. Mod. Phys. 81, 591–646 (2009) 6. S. Galam, B. Chopard, A. Masselot and M. Droz, “Competing Species Dynamics”, Eur. Phys. J. B 4, 529 (1998) 7. B. Chopard, M. Droz and S. Galam, “An Evolution Theory in Finite Size Systems”, Eur. Phys. J. B 16, Rapid Note, 575 (2000) 8. S. Galam, B. Chopard and M. Droz, “Killer geometries in competing species dynamics”, Physica A 314, 256 (2002) 9. J. Majewski, H. Li and J. Ott, “The Ising Model in Physics and Statistical Genetics”, The American Journal of Human Genetics 69, 853 (2001) 10. S. Galam, “The September 11 attack: A percolation of individual passive support”, Eur. Phys. J. B 26 Rapid Note, 269–272 (2002) 11. S. Galam and A. Mauger, “On reducing terrorism power: a hint from physics”, Physica A 323, 695–704 (2003) 12. S. Moss de Oliveira, P. M. C. de Oliveira, and D. Stauffer, “Evolution, Money, War, and Computers—Non-Traditional Applications of Computational Statistical Physics”, Teubner, Stuttgart-Leipzig (1999) References 89
13. W. Weidlich, “Sociodynamics; A Systematic Approach to Mathematical Modelling in the Social Sciences”, Harwood Academic Publishers, Amsterdam (2000) 14. Frank Schweitzer, Brownian Agents and Active Particles: On the Emergence of Complex Behavior in the Natural and Social Sciences, Springer, Berlin (2003) 15. D. Stauffer, S. Moss de Oliveira, P.M.C. de Oliveira, J.S. Sa Martins, “Biology, Sociology, Geology by Computational Physicists”, Elsevier, Amsterdam (2006) 16. B. K. Chakrabarti, A. Chakraborti, A. Chatterjee (Eds.), “Econophysics and Sociophysics: Trends and Perspectives”, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim (2006) 17. S. Galam, “Sociophysics: a personal testimony”, Physica A 336, 49 (2004) 18. K. G. Wilson and J. Kogut, “The renormalization group and the epsilon expansion”, Phys. Rep. 12C, 75 (1974) 19. K. G. Wilson and M. E. Fisher, “Critical Exponents in 3.99 Dimensions”, Phys. Rev. Lett. 28, 240 (1972) 20. S. Galam, “About imperialism of physics”, Fundamenta Scientiae 3, 125 (1982) 21. P. Pfeuty and S. Galam, “Les physiciens et la frustrations des electrons”,´ La Recherche. July–August, 23 (1981) 22. S. Galam and P. Pfeuty, “Physicists are frustrated”, Physics Today Letter, April, 89 (1982) 23. P. Ball, “Utopia theory”, Phys. World October, 7 (2003) 24. D. Stauffer, “Introduction to statistical physics outside physics”, Physica A 336, 1 (2004) 25. T. C. Schelling, “Dynamic Models of Segregation”, J. Math. Sociology 1, 143 (1971) 26. D. Stauffer and S. Solomon, “Ising, Schelling and self-organising segregation”, Eur. Phys. J. B 57, 473 (2007) 27. D. Stauffer and S. Solomon, “Applications of Physics and Mathematics to Social Science”, arXiv:0801.0121 (2008) 28. J. J. Schneider, private communication (2007) 29. S. Galam, Y. Gefen and Y. Shapir, Sociophysics: A mean behavior model for the process of strike, Journal of Mathematical Sociology 9, 1 (1982) 30. W. Weidlich, “The statistical description of polarization phenomena in society”, British. J. Math. Stat. Psychol. 24, 251 (1971) 31. W. Weidlich, “Synergetic modelling concepts for sociodynamics with application to collective political formation”, J. Math. Sociology, 18 (1994) 267–291 32. H. Wilhelmsson and L. Stenflo, “Nonlinearities and soliton-like structure in society”, Specula- tions in Science and technology 4, 297 (1981) 33. S. Galam, “Majority rule, hierarchical structures and democratic totalitarianism: a statistical approach”, J. Math. Psychology 30 426 (1986) 34. S. Galam, “Social paradoxes of majority rule voting and renormalization group”, J. Stat. Phys. 61, 943 (1990) 35. S. Galam, “Political paradoxes of majority rule voting and hierarchical systems”, Int. J. General Systems 18, 191 (1991) 36. S. Galam, “Real space renormalization group and social paradoxes in hierarchical organisa- tions”, in: Models of Self-organizationin Complex Systems (Moses), Akademie-Verlag, Berlin, vol. 64, 53 (1991) 37. S. Galam, Paradoxes de la regle majoritaire dans les systemes hirarchiques, Revue de Bibliologie 38, 62 (1993) 38. S. Galam, “Application of Statistical Physics to Politics”, Physica A 274, 132 (1999) 39. S. Galam, “Contrarian deterministic effect: the hung elections scenario”, Physica A 333, 453 (2004) 40. C. Borghesi and S. Galam, “Chaotic, staggered, and polarized dynamics in opinion forming: The contrarian effect”, Phys. Rev. E 73 066118 (2006) 41. S. Galam, “From 2000 Bush–Gore to 2006 Italian elections: voting at fifty–fifty and the contrarian effect”, Quality and Quantity Journal 41, 579 (2007) 42. K. Sznajd-Weron and J. Sznajd, “Opinion evolution in closed community”, Int. J. Mod. Phys. C 11, 1157 (2000) 90 4 Sociophysics: Weaknesses, Achievements, and Challenges
43. A.T. Bernardes, D. Stauffer and J. Kertsz, “Election results and the Sznajd model on Barabasi network”, Eur. Phys. J. B 25, 123 (2002) 44. S. Galam, “Fragmentation versus stability in bimodal coalitions”, Physica A 230, 174 (1996) 45. R. Florian and S. Galam, “Optimizing Conflicts in the Formation of Strategic Alliances”, Eur. Phys.J.B16, 189 (2000) 46. M. C. Gonzalez, A. O. Sousa and H. J. Herrmann, “Opinion formation on a deterministic pseudo-fractal network”, Int. J. Mod. Phys. C 15, 45 (2004) 47. S. Fortunato and C. Castellano, “Scaling and Universality in Proportional Elections”, Phys. Rev. Lett. 99, 138701 (2007) 48. S. Galam, “Le dangereux seuil critique du FN”, Le Monde, Vendredi 30 Mai, 17 (1997) 49. S. Galam, “Crier, mais pourquoi”, Liberation,´ Vendredi 17 Avril, 6 (1998) 50. J. J. Schneider and C. Hirtreiter, “The Impact of election results on the member numbers of the large parties in bavaria and germany”, Int. J. Mod. Phys. C 16 (2005) 1165–1215 51. J. J. Schneider and C. Hirtreiter, “Investigation of Election Results, Numbers of Party Members, and Opinion Polls in Germany”, Int. J. Mod. Phys. C 19, 441 (2008) 52. S. Galam, “Minority Opinion Spreading in Random Geometry”, Eur. Phys. J. B Rapid Note 25, 403 (2002) 53. S. Galam, “The dynamics of minority opinion in democratic debate”, Physica A 336, 56 (2004) 54. S. Galam, “Heterogeneous beliefs, segregation, and extremism in the making of public opinions”, Phys. Rev. E 71, 046123 (2005) 55. P. Lehir, “Les mathematiques´ s’invitent dans le debat´ europeen”,´ Le Monde, Samedi 26 Fevrier,´ 23(2005) 56. S. Galam, “Pourquoi des elections´ si serrees´ ?”, Le Monde, Mercredi 20 Septembre, 22 (2006) 57. S. Galam and F. Jacobs, “The role of inflexible minorities in the breaking of democratic opinion dynamics”, Physica A 381, 366 (2007) 58. S. Galam, “Pas de certitude scientifique sur le climat,”, Le Monde, Mercredi 07 Fvrier, 20 (2007) 59. V. Maurus, “Her´ esie”,´ Le Monde, Samedi 18 Fvrier, Chronique de la mediatrice´ (2007) 60. S. Galam, “Global warming : a social phenomena”, Complexity and Security, The NATO Science for Peace and Security Programme, Chapitre 13, J. J. Ramsden and P.J. Kervalishvili (Eds.) (2008) 61. S. Galam, Les scientifiques ont perdu le Nord, Reflexions´ sur le rechauffement´ climatique, Editions´ Plons, Paris (2008) Part II Discovering the Wonderful (and Maybe Scary) World of Sociophysics Chapter 5 Sociophysics: An Overview of Emblematic Founding Models
In this chapter, I review the series of rather different models that I have been developing over the past three decades (1980–2010). They go in several directions, and focus on five different sociopolitical phenomena. To have these subjects accepted by the physics community was a long and tedious fight [1]. The first one concerns the process of group decision making. It looks for eventual universal features behind all cases of group decision making. It may apply to small as well as to large groups and includes a large spectrum of situations such as a jury having to decide on the guilt of somebody accused of murder or workers who have to undergo a strike or even a military committee who have to choose between launching a minor or a major operation. Unexpected and deranging conclusions are obtained. The second focus is on the effects of bottom-up democratic voting in hierarchical systems. The emphasis is on the study of apparent natural biases. It shows how dictatorship mechanisms are driven by common sense rules. It provides a surprising explanation to last century’s sudden collapse of Eastern European communist parties. The third theme deals with changing the global state of a given system, which may be a society or a social group of agents. The question is to determine when and how the property of a local item succeeds in modifying the global property of the whole system. One illustration is the effect of having passive supporters for a terrorist group on its geographical capacity to act. Another illustration explains why opposite feelings can coexist within the same territory and both be legitimate, such as with the feelings of security and insecurity. The fourth item is related to the spontaneous formation of coalitions in society among people, enterprises, or countries. The same scheme applies for increasing size units. It analyzes both its stability as well as the corresponding dynamics of fragmentation. It provides an unusual view of Western European stability compared with the situation in Eastern Europe following the Warsaw pact collapse. The case of ex-Yugoslavia is studied in detail. New light is shed on the stability of the world during the cold war together with the role of coalitions.
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 93 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 5, © Springer Science+Business Media, LLC 2012 94 5 Sociophysics: An Overview of Emblematic Founding Models
The last point on the agenda articulates around the key issue of opinion dynamics, which has become a central element of democratic societies. Nowadays, most issues are discussed extensively among the public and institutional decisions have often complied to take into account the corresponding public opinion. On crucial decisions, winning or losing public opinion can be instrumental in being able to implement new regulations to impose behavior changes such as the right to smoke or not. To keep within the book’s pedagogical approach for each class of models, I briefly outline the original physics model and the techniques which have been used to construct the associated sociophysics framework. I also single out the main similarities as well as the salient differences between the physics model and its sociophysics counterpart. A series of snapshots of the numerous novel and counterintuitive results obtained with respect to the sociopolitical realities is included. It gives a taste of the rich variety and the drastically different angles of viewpoints with which to tackle them. An emphasis is put on the fact that using these models, several major real political events were successfully predicted within the first few years of the 21st century. It includes the prediction of the two totally improbable victories of, on the one hand, the French extreme right party in the 2000 presidential elections where it reached the second round against all odds, and the other, the “NO” vote to the 2005 French referendum on the European constitution, which was indeed unconceivable. Fifty– fifty elections were also successfully predicted to occur as they did so in several democratic countries including Germany, Italy, and Mexico. This chapter is an overview and not an in-depth investigation. For further details, a reading of the original research papers is required. The corresponding list is given in the references. On the contrary, the rest of the book focuses in great detail on the second item concerning bottom-up democratic voting and the establishment of de facto dictatorships. Each time a model is presented, the precise overlap with the physics model from which it was inspired, is outlined. Eventual novelties in the model are emphasized. The various results obtained from each class of models are enumerated, enlight- ening their novel and counterintuitive aspects with respect to the associated social and political frameworks. Among others, several major real political events were successfully predicted. These include the victory of the French extreme right party in the 2000 first round of the French presidential election [2–7], the voting at fifty– fifty in several democratic countries (Germany, Italy, Mexico) [8], and the victory of the “NO” vote in the 2005 French referendum on the European constitution [9–11]. For a survey of other work and reviews, I refer the reader to the following selection of references [12–14]. 5.1 In a Few Words 95
5.1 In a Few Words
At the beginning of the 21st century, sociophysics has become a recognized field of statistical physics. In the last ten years, hundreds of papers have been printed in leading international physics journals and their number is growing at a steady rate. In addition, quite a few international conferences and workshops are held every year. It is a flourishing and expanding field of research. Nevertheless, it is still anchored almost exclusively within physics. Being the opposite of the present situation, it is worth reminding that during the 1970s when sociophysics made its first faltering steps, it was received with hostility from the physics community. Only in the mid 1990s did it begin to attract several physicists from around the world. Then, 15 years were necessary to get it established up and running. Its next challenge is to build a bridge toward the social sciences, which is a different task from trying to bridge the gap with the social sciences. Sociophysics covers a growing number of topics. Numerous problems are addressed including voting, coalition formation, opinion dynamics, social networks, the evolution of languages, population dynamics, the spreading of epidemics, and terrorism. Among these topics, the subject of opinion dynamics and social networks have become mainstream in sociophysics, producing a great deal of research papers. I present here a review of the models I have been developing over the 30 years from 1980 to 2010. Most of them make up much of the pioneering work of sociophysics. The fact that they are not always given this credit results from the various flaws of the citation dynamics as well as the personal ambition of certain leading researchers. Hopefully, with time, increasing awareness, and some fighting, things will become more ethical. My models focus in particular on five sociopolitical issues including decision making, democratic voting in bottom-up hierarchical systems, and fragmentation versus coalitions among countries, opinion dynamics, and terrorism.
5.1.1 Decision Making
The question of decision making is central in many social organizations, whatever the level of their complexity. It occurs everywhere in various frameworks, right down to firms and small committees. Using the ferromagnetic Ising spin Hamil- tonian with both external and random-quenched fields at both zero and nonzero temperatures has proven very successful in tackling the problem [15–24]. In each case, the phase diagrams are constructed to obtain the landscape of the various decisions with the eventual possibility of determining the final decision. The effect of reversing an external field on the collective equilibrium state has been studied with an emphasis on the existence of nucleation phenomena. Mean field treatment is applied. The inclusion of individual biases has also been studied. 96 5 Sociophysics: An Overview of Emblematic Founding Models
5.1.2 Bottom-up Democratic Voting
Another crucial question concerns the procedure for designating representatives in order to decide and carry out policy. Often, in democratic countries, but also in nondemocratic countries, bottom-up hierarchies operate so as to asses that goal. I investigate throughout the paradox of using bottom-up democratic voting to establish de facto a dictatorship. But it is not a repressive dictatorship, only one that uses repeated voting. The possible twisting of the democratic character of using majority rules is brought to light as well as the limits of the manipulation [25–37]. The models consider a mixed population with two groups of agents A and B. A is then built from the population using local majority rules with the possibility of some power inertia bias. Tree-like networks are thus constructed, which combine a random selection of agents at the bottom from the surrounding population with an associated deterministic outcome at the top. The scheme relates to the adapting of real space renormalization group techniques in building a social and political structure.
5.1.3 Terrorism
A rather puzzling question concerns the understanding of the connection between some local properties attached to a person and the appearance or not, at the global level of the corresponding social group, of a novel property. In the case in which such a property emerges, it is usually almost impossible to erase it since there exists no direct link between it and a given individual sharing the property. The concept of percolation is found to be quite adequate to tackle this intriguing feature of social systems. It has been used to provide a consistent explanation of how it is possible to have simultaneously a feeling of safety and a feeling of a lack of safety on the same territory, in which both feelings are indeed legitimate [5]. The effect of having passive supporters of a terrorist group has also been investigated through the prism of percolation. It determines the actual geographic extension of the territory, which is open to terrorists for free movement [38–43].
5.1.4 Coalitions Versus Fragmentation
Once the Warsaw pact was disbanded, I was quite unconvinced of the rationale proposed to explain such a historical event. In particular I disagreed with the arguments given to justify the various civil wars which burst out in Eastern Europe as well as the arguments given to justify the Western European stability. On this basis, a combination of random bond and random site spin glasses was found to References 97 provide a powerful framework with which to describe the formation of coalitions and the dynamics of fragmentation among a group of countries. External and local fields were also considered together with site dilution effects in mixtures of ferro- and antiferromagnetic spin Hamiltonians. Ising and Potts variables have been used. A consistent explanation of the European situation was obtained together with some detailed analysis of the wars, which teared the ex-Yugoslavia apart [44–48].
5.1.5 Public Opinion
Public opinion has become a sensitive ingredient in the now overconnected global world, and in particular with the spread of the use of the internet. Even nondemo- cratic countries are now having to face up to public opinion. One most salient feature is the minority spreading effect. How and why do minority opinions, usually either against changes or in support of prejudiced views, have the final favor of large groups of people, which discuss the issue freely? The question is solved by constructing reaction–diffusion like models which combine local majority rules and agent reshuffling. Several kinds of agents are also considered. These include floaters, agents that follow the local majority, contrarians, who oppose the majority, either local or global, and inflexibles, who never change their mind [49–64]. The competition between two and three opinions is analyzed. Techniques from the real space renormalization group approach are used to solve the equations.
References
1. S. Galam, “Sociophysics: a personal testimony”, Physica A 336, 49–55 (2004) 2. S. Galam, “Le dangereux seuil critique du FN”, Le Monde, Vendredi 30 Mai, 17 (1997) 3. S. Galam, “Crier, mais pourquoi”, Liberation,´ Vendredi 17 Avril, 6 (1998) 4. S. Galam, “Le vote majoritaire est-il totalitaire ?”, Pour La Science, Hors serie,´ Les Mathematiques´ Sociales, 90–94 July (1999) 5. S. Galam, “Citation” in an editorial from Jean dOrmesson, front page, daily newspaper Le Figaro, Mardi 4 Juin, 1 (2002) 6. S. Galam, “Risque de raz-de-maree´ FN”, Entretien, France Soir, La Une et 3, Mercredi 5 Juin (2002) 7. S. Galam, “Le FN au microscope”, Le Minotaure 6, 88–91, Avril (2004) 8. S. Galam, “Pourquoi des elections´ si serrees´ ?”, Le Monde, Mercredi 20 Septembre, 22 (2006) 9. S. Galam, “Les mathematiques´ s’invitent dans le debat´ europeen”,´ Interview par P. Lehir, Le Monde, Samedi 26 Fevrier,´ 23 (2005) 10. S. Galam, “Les mathematiques´ s’invitent dans le debat´ europeen”,´ Le Monde, Lundi 11 Avril, 15 (2005), Reproduced in the international weekly selection of Le Monde 2005 11. S. Galam, “Les mathematiques´ s’invitent dans le debat´ europeen”,´ Le Monde, Lundi 11 Avril, 15 (2005), Reproduced in “TA NEA”, Greek daily newspaper March 3 (2005) 12. Stauffer, D., S. Moss de Oliveira, P. de Oliveira, and J. Sa Martins, “Biology, sociology, geology by computational physicists”, Elsevier, Amsterdam (2006) 98 5 Sociophysics: An Overview of Emblematic Founding Models
13. Econophysics and Sociophysics: Trends and Perspectives, B. K. Chakrabarti, A. Chakraborti, A. Chatterjee (Eds.), Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim (2006) 14. C. Castellano, S. Fortunato and V. Loreto, “Statistical physics of social dynamics”, Rev. Mod. Phys. 81, 591–646 (2009) 15. S. Galam, Y. Gefen and Y. Shapir, “Sociophysics: A mean behavior model for the process of strike”, Math. J. of Sociology 9, 1–13 (1982) 16. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: Consensus and attitude changes in groups”, Euro. J. of Social Psy. 21, 49–74 (1991) 17. S. Galam and S. Moscovici, “Compromise versus polarization in group decision making”, in Defense Decision Making, R. Avenhaus, H. Karkar and M. Rudnianski (Eds), Springer-Verlag, Berlin, 40–51 (1991) 18. S. Galam and S. Moscovici, “A theory of collective decision making in hierarchical and non- hierarchical groups”, Russian Psy. J. 13, 93–103 (1993) 19. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: II. Conformity and power”, Euro. J. of Social Psy., 24, 481–495 (1994) 20. S. Galam and S. Moscovici, “Towards a theory of collective phenomena: III. Conflicts and forms of power”, Euro. J. of Social Psy., 25, 217–229 (1995) 21. S. Galam, “When humans interact like atoms”, Understanding group behavior, vol. I, Chap. 12, 293–312, Davis and Witte, Eds, Lawrence Erlbaum Ass., New Jersey (1996) 22. S. Galam, “Rational group decision making: a random field Ising model at T D 0”, Physica A, 238, 66–80 (1997) 23. S. Galam and J. D. Zucker, “From Individual Choice to Group Decision Making”, Physica A 287, 644–659 (2000) 24. S. Galam, “Universality of Group Decision Making”, Traffic and Granular Flow ’99 D. Helbing et al, Eds., Springer, Berlin (2000) 25. S. Galam, “Majority rule, hierarchical structures and democratic totalitarism: a statistical approach”, J. of Math. Psychology 30, 426–434 (1986) 26. S. Galam, “Social paradoxes of majority rule voting and renormalization group”, J. of Stat. Phys. 61, 943–951 (1990) 27. S. Galam, “Political paradoxes of majority rule voting and hierarchical systems”, Int. J. General Systems 18, 191–200 (1991) 28. S. Galam, “Real space renormalization group and social paradoxes in hierarchical organi- sations”, Models of self-organization in complex systems (Moses) Akademie-Verlag, Berlin V. 64, 53–59 (1991) 29. S. Galam, “Paradoxes de la regle majoritaire dans les systemes hierarchiques”,´ Revue de Bibliologie, 38, 62–68 (1993) 30. S. Galam, “Application of Statistical Physics to Politics”, Physica A 274, 132–139 (1999) 31. S. Galam, “Real space renormalization group and totalitarian paradox of majority rule voting”, Physica A 285, 66–76 (2000) 32. S. Galam and S. Wonczak, “Dictatorship from Majority Rule Voting”, Eur. Phys. J. B 18, 183–186 (2000) 33. S. Galam, “Democratic Voting in Hierarchical Structures”, Application of Simulation to Social Sciences, G. Ballot and G. Weisbush, Eds. Hermes, Paris, 171–180 (2000) 34. S. Galam, “Building a Dictatorship from Majority Rule Voting”, ECAI 2000 Modelling Artificial Societies, C. Jonker et al, Eds., Humboldt U. Press (ISSN: 0863-0957), 23–26 (2001) 35. S. Galam, “How to Become a Dictator”, Scaling and disordered systems. International Workshop and Collection of Articles Honoring Professor Antonio Coniglio on the Occasion of his 60th Birthday. F. Family. M. Daoud. H.J. Herrmann and H.E. Stanley, Eds., World Scientific, 243–249 (2002) 36. S. Galam, “Dictatorship effect of the majority rule voting in hierarchical systems”, Self- Organisation and Evolution of Social Systems, Chap. 8, Cambridge University Press, C. Hemelrijk (Ed.) (2005) 37. S. Galam, “Stability of leadership in bottom-up hierarchical organizations”, Journal of Social Complexity 2 62–75 (2006) References 99
38. S. Galam, “The September 11 attack: A percolation of individual passive support”, Eur. Phys. J. B 26 Rapid Note, 269–272 (2002) 39. S. Galam and A. Mauger, “On reducing terrorism power: a hint from physics”, Physica A 323, 695–704 (2003) 40. S. Galam, “Global physics: from percolation to terrorism,: guerilla warfare and clandestine activities”, Physica A 330, 139–149 (2003) 41. S. Galam, “La detection´ des reseaux´ terroristes”, Strategie´ et decision´ : La crise du 11 septembre, General Loup Francart et Isabelle Dufour, Economica, Paris (2002) 42. S. Galam, “Terrorisme et percolation”, Pour La Science 306, 90–93, Avril (2003) 43. S. Galam, “Global terrorism versus social permeability to underground activities”, in Econophysics and Sociophysics: Trends and Perspectives, B. K. Chakrabarti, A. Chakraborti, A. Chatterjee (Eds.), Chap. 14, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim (2006) 44. S. Galam, “Comment on A landscape theory of aggregation”, British J. Political Sciences 28, 411–412 (1998) 45. S. Galam, “Fragmentation versus stability in bimodal coalitions”, Physica A 230, 174–188 (1996) 46. S. Galam, “Spontaneous coalitions forming: a model from spin glass”, arXiv:cond- mat/9901022 (1999) 47. R. Florian and S. Galam, “Optimizing Conflicts in the Formation of Strategic Alliances”, Eur. Phys. J. B 16, 189 (2000) 48. S. Galam, “Spontaneous Coalition Forming. Why Some Are Stable?”, Lecture Notes in Computer Science. 5th International Conference on Cellular Automata for Research and Industry, Springer-Verlag Heidelberg, 1–9 Vol. 2493 (2002) 49. S. Galam, B. Chopard, A. Masselot and M. Droz, “Competing Species Dynamics”, Eur. Phys. J. B 4, 529–531 (1998) 50. B. Chopard, M. Droz and S. Galam, “An Evolution Theory in Finite Size Systems”, Eur. Phys. J. B 16, Rapid Note, 575–578 (2000) 51. S. Galam, “Minority Opinion Spreading in Random Geometry”, Eur. Phys. J. B 25 Rapid Note, 403–406 (2002) 52. S. Galam, B. Chopard and M. Droz, “Killer geometries in competing species dynamics”, Physica A 314, 256–263 (2002) 53. S. Galam, “Modeling Rumors: The No Plane Pentagon French Hoax Case”, Physica A 320, 571–580 (2003) 54. S. Galam, “Contrarian deterministic effect: the hung elections scenario”, Physica A 333, 453–460 (2004) 55. S. Galam, “The dynamics of minority opinion in democratic debate” Physica A 336, 56–62 (2004) 56. S. Galam and Bastien Chopard, “Threshold Phenomena versus Killer Clusters in Bimodal Competion for Standards”, Cognitive Economics – An Interdisciplinary Approach, P. Bourgine and J.-P. Nadal, Eds, Springer, 429–440 (2004) 57. S. Galam and A. Vignes, “Fashion, novelty and optimality: an application from Physics”, Physica A 351, 605–619 (2005) 58. S. Gekle, L. Peliti, and S. Galam, “Opinion dynamics in a three-choice system”, Eur. Phys. J. B 45, 569–575 (2005) 59. S. Galam, “Heterogeneous beliefs, segregation, and extremism in the making of public opinions”, Phys. Rev. E 71, 046123-1-5 (2005) 60. S. Galam, “Local dynamics vs. social mechanisms: A unifying frame”, Europhys. Lett. 70, 705–711 (2005) 61. C. Borghesi and S. Galam, “Chaotic, staggered, and polarized dynamics in opinion forming: The contrarian effect”, Phys. Rev. E 73 066118 (1–9) (2006) 62. S. Galam, “Opinion dynamics, minority spreading and heterogeneous beliefs”, in Econo- physics and Sociophysics: Trends and Perspectives, B. K. Chakrabarti, A. Chakraborti, A. Chatterjee (Eds.), Chap. 13, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim (2006) 100 5 Sociophysics: An Overview of Emblematic Founding Models
63. S. Galam, “From 2000 Bush–Gore to 2006 Italian elections: voting at fifty–fifty and the contrarian effect”, Quality and Quantity Journal 41 579–589 (2007) 64. S. Galam and F. Jacobs, “The role of inflexible minorities in the breaking of democratic opinion dynamics”, Physica A 381 366–376 (2007) Chapter 6 Universal Features of Group Decision Making
Anyone familiar with the so-called Ising ferromagnetic model of modern statistical physics would certainly be tempted to make a connection with the behavior of people. Its major feature relies on its incredible universality. Many situations in the world are governed by the emergence of a global order from local interactions. It has proven very powerful in explaining the properties of many different physical systems. It is therefore a very appealing universal model, which could certainly also apply to a large spectrum of social situations. Any book on statistical mechanics will provide an introduction to the Ising model. Among them, I recommend the books by Pathria, Reif, and Ma [1–3].
6.1 The Strike Phenomenon
In the late 1970s, many people had leftist visions of society and it being reworked, making the strike a central key to implementing economic changes in the organiza- tion of society. Within this frame of mind, I was puzzled by the fact that sometimes a tiny group of activists succeeded in putting on strike thousands of workers while at other times even large unions could not achieve this. Discussing the question with Y. Shapir and Y. Gefen [4], both PhD. students like me in the same group at the Department of Physics at Tel Aviv University, we devel- oped the idea of using an Ising ferromagnetic system to describe the collective state of an assembly of agents, each being in either one of two distinct individual states, that of working or striking. This produces two collective ordered states: a working state versus a striking state. The ferromagnetic coupling between agents was moti- vated by the social fact that people have the tendency to reproduce the leading choice of their neighbors, in particular in conflicting situations. We thus implemented the first application of the Ising model to describe the global state of a firm.
S. Galam, Sociophysics: A Physicist’s Modeling of Psycho-political Phenomena, 101 Understanding Complex Systems, DOI 10.1007/978-1-4614-2032-3 6, © Springer Science+Business Media, LLC 2012 102 6 Universal Features of Group Decision Making
6.1.1 The Model
A group of N agents is considered, with each one having attached to it a variable SQi where i D 1;2;:::;N. When the agent i is working at the maximum of its productivity, we take SQi D 1. In the opposite case, when it is not working at all, we put SQi D 0. Instead of the two values 0 and 1, we need to have ˙ to fit the Ising model. This is readily achieved by making the variable change Si D 2 SQi 1=2 . Now a maximum productivity by agent i is denoted by Si D 1 while a zero activity is defined by Si D 1. Following the Ising model construction, we introduce a coupling Ji;j between two agents i and j . This coupling embodies their social exchanges. As in ferromag- netism, we take a positive coupling with Ji;j >0. The quantity Ji;j Si Sj is then a measurement of the discrepancy which exists between their individual states. This accounts for an eventual dissatisfaction generated by a difference in productivity. Otherwise, when they are synchronized, a negative dissatisfaction is obtained, which means that the satisfactions are alike. To account for the dissatisfaction created by the actual salary W with respect to an expected minimum fair salary expectation V , we introduce an external field denoted H D W V , which couples linearly with each agent to create a dissatisfaction HSi . Accordingly, when H>0, if the agent is working at maximum, it gets the maximum negative dissatisfaction H, i.e., a satisfaction while H<0, to reach the maximum negative dissatisfaction H, the agent must stop working. The first case is activated by having W>Vand the second by W X XN H D Ji;j Si Sj H Si ; (6.1) .i;j / iD1 where (i, j) denotes all interacting pairs of agents. 6.1.2 The Operating Mechanism We are now in a position to be able to postulate a principle for determining which state, working or not working, a given agent will select, given a distribution of the individual states of all the other agents. A reasonable postulate is to assume 6.1 The Strike Phenomenon 103 Fig. 6.1 Dissatisfaction function F versus M in the ordered phases. The cross x denotes the actual state of the system. It is stable since it is at the absolute minimum. Here H>0and the factory is at work that an agent would like to minimize its dissatisfaction. A principle of minimum dissatisfaction is thus postulated to determine the eventual equilibrium global state of the system. It is analogous to the minimum energy principle which is activated at a temperature of absolute zero. In physics, the temperature T is a crucial ingredient that does not exist as such in social systems. However, the notion of a social permeability denoted 1=T might be a good counterpart. The next step is then to introduce a global dissatisfaction function F , the equivalent of the physical free energy. Formally, it can be calculated exactly, since it is an explicit function of H and T . However, in practice, for most of the cases, this can only be performed by numerical means. Approximate techniques exist to perform some of the analytical calculations. The most powerful one is called the mean field treatment since it neglects local fluctuations. Accordingly, we performed a mean field treatment of (6.1), which quite naturally reproduces the well-known phase diagram of the ferromagnetic Ising model in an external field. A key internal parameter to characterize the macroscopic state of the system is the order parameter XN M 1=N Si : (6.2) iD1 A nonzero M reveals an ordered phase while M D 0 denotes the disordered phase. Several ordered phases might exist. In the present case, two are present with M>0and M<0. However, they are symmetrical, i.e., they display the same broken symmetry. An important property relevant to the strike phenomena is the existence of the metastability phenomenon combined with the nucleation mechanism, which are obtained by reversing the sign of the external field H. This situation occurs in particular when the expected minimum fair salary V increases in the workers expectation while the current wages are unchanged. The phenomenon is illustrated in the series of Figs. 6.1–6.4. 104 6 Universal Features of Group Decision Making Fig. 6.2 Dissatisfaction function F versus M in the ordered phases. The cross x denotes the actual state of the system. It is locally stable but could turn unstable since it is not at the absolute minimum. Here H<0and the factory is at work. The dashed arrow indicates the eventual jump into a strike state, which corresponds to the absolute minimum, driven by some external action such as the intervention of a handful of activists Fig. 6.3 The limit of metastability has been reached. We still have H<0but its magnitude has increased. F has only one minimum 6.1.3 The Overlap with the Physical Model The social model of strikes proposed here is identical to the Ising ferromagnet in an external field. The treatment used to obtain the corresponding phase diagram is the classical mean field theory. The totality of the associated rich properties are thus recovered and must be given a social interpretation. It is indeed a mapping from a physical reality to a social reality with all the limitations that such reasoning implies. 6.1 The Strike Phenomenon 105 Fig. 6.4 M versus H for K 6.1.4 The Novel Counterintuitive Social Highlights If technically the model does not bring anything new, the transfer from statistical physics to the new field of social sciences sheds new light on it. It produces a rather large spectrum of novel insights that are rather counterintuitive and unexpected. Most salient are [4]: • From our individual two-state assumption for each agent, the two collective states of a firm, being working normally or on strike, appear to be symmetrically ordered phases. Accordingly, the worker’s involvement in an eventual strike will be identical to the prevailing amplitude of working just before the strike begins. • It appears from the hysteresis phenomenon, which is linked to metastability, that avoiding a strike is not equivalent to stopping a strike in terms of cost. Having a positive effective wage H>0guarantees a working state with a production value M>0. However, the wage situation may well-deteriorate even if the actual salary is kept unchanged. It might happen when the worker’s expectation of the minimum salary E increases while no increase in the current salary is given. Such a change induces a field modification from H>0to H<0. In this case, although H<0, the firm’s working state can persist with no strike occurring, creating the illusion that everything is alright while indeed the system has moved from a stable state to a metastable state. Only below the limit of metastability, i.e., a more negative value of H does the strike happen with certainty. • However, while within the metastable working state, any handful of activists, including external ones, with very little action, could suddenly precipitate the whole firm into striking. The same action would have no effect in the range H>0. This is the manifestation of the so-called nucleation phenomenon, which 106 6 Universal Features of Group Decision Making takes place within the H range for which the striking phase corresponds to the absolute equilibrium and the working state is only a local equilibrium. • A remarkable feature is that our model provides us with an explanation of how and why it is often cheaper to avoid a strike by increasing W to reach a H>0 before the strike starts than to put a firm back to work once the strike has occurred. It is also the direct result of the existence of a hysteresis effect produced by the metastability. We can now have H>0and still have the workers staying on strike. This is the symmetrical situation of having them working when H<0. However, the major difference is that in the case of H<0, activists may well precipitate the firm into striking, that will not be the case when H>0. Then the minimum increase of the actual salary W to get the workers back to work is that which is necessary to get out of the metastable range as seen in Fig. 6.4. • In the case where the social permeability is weak enough to put the firm in the disordered state characterized by M D 0, the metastability vanishes. Then the current wage is instrumental in keeping the firm working or striking. As soon as H<0the agents strike and as soon as H>0, they work. The concern might be that M D 0 means an average individual production equal to zero. The workers are not much involved in their work with a low productivity. 6.1.5 Achievements of the Model While the model provides a new view with a series of guidelines on how to manage wage policy and how to implement a strike, no specific predictions have been made to date concerning a precise situation. 6.2 How Do Groups Make Their Decisions? The major contribution from the preceding model has been to initiate and to open up a new field of knowledge, taking over 30 years to attract a substantial part of the community of condensed matter physicists. These results in itself show the new insight that can be gained from sociophysics. Following the development of this model, I met in New York a leading social psychologist, Serge Moscovici. While discussing the possible applications of sociophysics, we decided to address the challenge of explaining the observed polarization and risk taken in a series of experiments conducted in psychology. This work led to the publication of a series of papers, which created a coherent and rather rich framework [5–9]. Indeed, it was found that groups that have no time limit constraint exhibit a trend towards an extreme decision instead of a balanced decision [10, 11]. As for the striking phenomenon, the Ising ferromagnetic model in an external field was perfectly adapted to the problem of a two-choice situation. However, while 6.2 How Do Groups Make Their Decisions? 107 taking again the same formal Hamiltonian given by (6.1), the parameters are given different meanings besides that of the coupling Ji;j , which keeps on measuring the strength of exchanges between two individuals, denoted i and j . After a series of very long in-depth discussions with Serge Moscovici, I concluded that using a modified version of a finite size of the so-called random field Ising ferromagnetic model in an external magnetic field at zero temperature would be perfectly adapted to the investigation of the social phenomenon of group decision making. Postulating a minimum of interpersonal conflicts, we obtained the psychosociological experimental fact that interactions produce a group polarization along a single choice, which is, however, found to be randomly selected in our approach. On this basis, a small external social pressure has a radical effect on the polarization. Individual bias related to personal backgrounds, cultural values, and past experiences can also be introduced via quenched local competing fields. These biases are shown to be instrumental in generating a larger spectrum of collective new choices beyond the initial ones, which in turn matches certain experimental findings. In particular, compromise is found to result from the existence of individual competing biases. Conflict is shown to weaken group polarization. The model yields new psychosociological insights about consensus and compromise in groups. To apprehend the numerous associated insights, we refer to the related papers for details [5–9, 12–15]. 6.2.1 The Symmetrical Individual Versus the Symmetrical Group In numerous cases, an individual has to make a choice between two given answers, which can be denoted by yes and no. In addition, a larger spectrum of answers can often be mapped at some approximate level into a two-answer case [16]. A variable ci D˙1 represents the choice of agents i with Yes ci DC1 and No ci D 1. At this stage, the individual has no reason to favor either one of the two choices, so the yes and no are equally probable with a probability distribution function 1 p.c / ı.c 1/ ı.c 1/ ; i D 2 f i C i C g (6.3) where ı is the Kronecker function. For N individuals, each one making its choice independently, the aggregated collective choice of a group of N people XN C D ci (6.4) iD1 p1 C 0 is thus zero on average, with fluctuations of the order of N .Theresult D creates a new qualitative choice which did not exist at the individual level. It can 108 6 Universal Features of Group Decision Making be understood as the perfect choice of compromise. Since the macroscopic quantity C is zero, the group aggregation process has no effect at the macrolevel, making the associated group symmetrical. It is a neutral group with no link among group members. The various choices of two people i and j are of two types. One type is where each makes an identical choice with ci D cj D˙1 and the other type is where they hold opposite choices with ci D cj D˙1. The product ci cj discriminates these two situations with ci cj DC1 for agreement and ci cj D 1 for conflict. The choice ˙1 is arbitrary, in accordance with both a symmetrical agent and a symmetrical group. This is the physical equivalent of time reversal invariance. The system is unchanged by reversing all individual spins. However, in a social system, although the individual has no a priori to select either one of the two choices, his or her actual choice will have a drastically different effect on the social implementation of the choice. Prior to the decision itself, although both choices are equally probable, both individuals could either argue for a long time or discuss and exchange information. On the other hand, they may well decide without any exchanges. A quantity is thus required to measure their choice of involvement. Calling J the exchange amplitude, the product Jci cj measures the amplitude of the pair involvement, either positive CJ , i.e., cooperation, or negative, meaning a conflict J . Restricting interaction to pairs, the overall group conflict is measured by the function X G J ci cj ; (6.5) where we have assumed a constant exchange J for all interacting pairs and represents all interacting pairs. G is the group internal conflict function. 6.2.2 The Random Symmetry Breaking Choice For each one of the 2N configurations of the N individual choices, the internal conflict function G measures the corresponding conflict. It discriminates between various possible choices but does not induce a choice. To produce a group decision dynamics, we need to introduce a criterion to select the actual collective choice favored by the group. In the spirit of the physical minimum energy principle, and from psychological evidence, we postulate that “Each individual favors the choice which minimizes his or her own conflict.” Minimum conflict means maximum agreement. At this stage agents make choices but are not personally bound to them. Later, we will introduce the personal convictions of agents. If we proceed in a sequential form, selecting randomly one person who made a choice, yes or no, then we select all people interacting with it and they will make the same choice so as to minimize their own conflicts. By so doing, all people 6.2 How Do Groups Make Their Decisions? 109 interacting with them will again make the same choice so as to favor agreement and so on and so forth. If none of the agents are isolated and if there exist no isolated subgroups, the group ends up making the same initial choice that the first person selected. The net result of these successive choices is an extreme polarization of the collective choice with C D˙N . The sign, i.e., the polarization direction, is determined by the initial individual choice, which was made randomly. This polarization phenomenon holds whoever is chosen to be the initial person. Only the direction (yes or no) will change from one initial person to another. In real life situations, the above process starts simultaneously from several people. The dynamics of choice spreading is a rather complex phenomenon. Monte- Carlo simulations at zero-temperature of the Ising model show nontrivial behavior at all dimensions [17]. However, the group eventually succeeds in selecting only one choice, allowing everyone to minimize their own conflicts. We thus find that, “Symmetrical groups polarize themselves towards an extreme choice. The direction of that choice is arbitrary. Each extreme is equally probable.” This shows that individual minimum conflict is identical to maximizing G.A positive conflict is an agreement. This polarization effect is identical to the physical spontaneous symmetry breaking phenomenon, which is instrumental in the making of collective phenomena in inert matter [18]. Individual local interactions make the group behave as a single super-person [19]. This super-person chooses between the two possible choices with equal probability in the same way as the isolated individual. Nevertheless, individuals within the group have lost their freedom of choice. Each individual must make a choice that is identical to that of the person with whom they interact. Individual freedom disappears in favor of the emergence of a group freedom. Perfect compromise has also disappeared. The group decision produces an effect on its surroundings that is somehow proportional to N since C D˙N . Without interactions, the individual choices ci D˙1 were overall self-neutralized macroscopically. Interactions produce strong individual correlations associated with the symmetry breaking. Results obtained from group decision making experiments conducted in social psychology can now be understood in a different way. The polarization effect was clearly evident in data reported in [16] but most theoretical explanations have been unconvincing in connecting choices at the individual level [20] and the group level [11]. We conclude from the model that the polarization effect arises quite naturally from first principles, i.e., from interactions among agents. 6.2.3 Anticipating the Group Choice Anticipation is a very crucial feature in most social phenomena. To incorporate it, we rewrite G from (6.5)as 110 6 Universal Features of Group Decision Making ( ) I XN Xn G c c ; D 2 j.k/ i (6.6) iD1 kD1 where n is the number of people one individual interacts with. To keep the presentation simple this number is assumed to be equal for everyone. For small groups where everyone interacts with everyone n D N . Anticipating the emergence of a collective choice, each agent i thus tries to project its expectation of the overall final group decision through the various choices cj of the agents with whom it discusses. Individual i extrapolates j ’s choice cj to the expected collective choice the group will eventually make without its own contribution. Accordingly individual i perceives j ’s choice as given by 1 cj ! .C ci /; (6.7) N 1 where C is the collective choice. Within this anticipation process, the function G is written 8 9 J XN