Book contents
- Frontmatter
- Dedication
- Contents
- List of Figures
- List of Tables
- Preface and Acknowledgments
- Introduction
- 1 Preference
- 2 Aggregation
- 3 Deliberation
- 4 Coordination
- 5 Randomization
- 6 Satisficing
- Appendix A Dutch Book Theorem
- Appendix B Bayesian Networks
- Appendix C Probability Concepts
- Appendix D Markov Convergence Theorem
- Appendix E Entropy and Mutual Information
- Bibliography
- List of Authors
- Index
Preface and Acknowledgments
Published online by Cambridge University Press: 05 September 2016
- Frontmatter
- Dedication
- Contents
- List of Figures
- List of Tables
- Preface and Acknowledgments
- Introduction
- 1 Preference
- 2 Aggregation
- 3 Deliberation
- 4 Coordination
- 5 Randomization
- 6 Satisficing
- Appendix A Dutch Book Theorem
- Appendix B Bayesian Networks
- Appendix C Probability Concepts
- Appendix D Markov Convergence Theorem
- Appendix E Entropy and Mutual Information
- Bibliography
- List of Authors
- Index
Summary
Network theory provides a powerful and expressive framework for the analysis and synthesis of collectives whose members exert social influence on each other. When such a collective is engaged in a social choice, all social relationships that could influence the decision must be taken into consideration. This book advances social choice theory by introducing extended concepts of preference, aggregation, deliberation, and coordination that enable the group to incorporate social influence relationships into a comprehensive social model from which a coordinated social choice can be deduced.
Historically, social choice theory has focused mainly on the study of human behavior and has principally fallen under the purview of the social sciences. Increasingly, however, computer science has applied social choice theory to the design and synthesis of artificial societies such as multiagent systems and networks. A principle motivation for this book is to present a view of the theory that is applicable to both cultures.
Although both the social science and computer science disciplines rely on abstract mathematical models, they use them differently. Social science uses social models primarily as analysis tools to understand, predict, explain, or recommend behavior for human society. Such models may provide useful insights regarding social behavior, but they are not causal – they do not dictate behavior. They are idealized approximations whose validity hinges on assumptions regarding human social behavior. Computer science and engineering, however, use social models as synthesis tools to design and construct artificial social systems populated by autonomous agents who are designed to function in ways that are compatible with human behavior. In this sense, the models are causal, since they generate the behavior of the members of the society as they interact.
The difference between analysis and synthesis is that with analysis, models are used to reduce reality to an abstraction, while synthesis uses models to create a reality from an abstraction. The difference between these two applications is important. With analysis, psychological or sociological attributes such as cooperation and altruism, or even such overtly antisocial attributes as conflict and avarice, can be ascribed to individuals as a function of the solution concept, even if such attributes are not formally part of the mathematical model. But when synthesizing an artificial society, such attributes must be explicitly incorporated into the mathematical model or they will not exist.
- Type
- Chapter
- Information
- Theory of Social Choice on NetworksPreference, Aggregation, and Coordination, pp. xv - xviiiPublisher: Cambridge University PressPrint publication year: 2016