Generative Social Science: Studies in Agent-Based Computational Modeling (Princeton Studies in Complexity) [Hardcover]

Joshua M. Epstein (Author)

Book Description

Publication Date: January 8, 2007 | ISBN-10: 0691125473 | ISBN-13: 978-0691125473

Agent-based computational modeling is changing the face of social science. In Generative Social Science, Joshua Epstein argues that this powerful, novel technique permits the social sciences to meet a fundamentally new standard of explanation, in which one "grows" the phenomenon of interest in an artificial society of interacting agents: heterogeneous, boundedly rational actors, represented as mathematical or software objects. After elaborating this notion of generative explanation in a pair of overarching foundational chapters, Epstein illustrates it with examples chosen from such far-flung fields as archaeology, civil conflict, the evolution of norms, epidemiology, retirement economics, spatial games, and organizational adaptation. In elegant chapter preludes, he explains how these widely diverse modeling studies support his sweeping case for generative explanation.

This book represents a powerful consolidation of Epstein's interdisciplinary research activities in the decade since the publication of his and Robert Axtell's landmark volume, Growing Artificial Societies. Beautifully illustrated, Generative Social Science includes a CD that contains animated movies of core model runs, and programs allowing users to easily change assumptions and explore models, making it an invaluable text for courses in modeling at all levels.

Editorial Reviews

Review

It should be noted that having all these contributions in one place is not only useful but pleasing...Epstein's book is a concise and well articulated defense of agent-based modeling. Generative Social Science is essential reading for anyone seriously interested in the foundations and the practice of agent-based modeling. -- Daniel Diermeier, Science

Epstein's Generative Social Science . . . is to be regarded as a success. It is a highly professional book, comestible also by non-experts without giving up scientific rigour. Probably because the author is fond of its subject matter, and manages to transfer his enthusiasm into the reader, the book may be read all at once, as a narrative. . . . In sum, there are good reasons to expect that the community of simulators will welcome this book with enthusiasm, and that other supporters will be recruited. -- Rosaria Conte, JASSS

Epstein's generative manifesto is essential reading for anyone seriously interested in explaining social life. -- Michael Macy, American Journal of Sociology

From the Inside Flap

"Joshua Epstein has been a leader in articulating and pursuing the agent-based generative approach to social science. This collection of his papers exemplifies both the depth of his methodological positions and the fruitfulness of agent-based analysis. The power of simple rules of local social interaction in generating explanations of complex social behavior is beautifully illustrated, most notably in the study of population fluctuations in Anasazi societies. I am convinced that agent-based approaches to economics will become a major tool."--Kenneth J. Arrow, Stanford University

"Generative Social Science is an outstanding example of an exciting paradigm shift in the analysis of dynamic social systems. Joshua Epstein is a virtuoso at using simple models to reveal surprising insights about the dynamics of a wide range of phenomena such as epidemics, status hierarchies, civil violence, and even the timing of retirement."--Robert Axelrod, University of Michigan

"Agent-based computational modeling represents an important new interdisciplinary approach to doing social science. Joshua Epstein, a pioneer of this approach, provides in Generative Social Science both a spirited defense of agent-based modeling and a dazzling display of the method's power."--John Duffy, University of Pittsburgh

"Epstein is a central and outstandingly creative figure in the emerging social science literature developed through agent-based simulation studies. Epstein offers an undogmatic, balanced account of his project and methods and shows in what specific ways they can open up whole broad questions that are simply unapproachable with traditional methods. The chapters address a stunningly wide range of problems, and each chapter has a distinctive and stimulating contribution to make."--Duncan K. Foley, New School for Social Research

"Generative Social Science is an important book that should be read by all who have a serious interest in the social sciences."--Peter Hedstrm, University of Oxford

"The contents are important, and until now have appeared in scattered and sometimes obscure places. The new commentary that the author has added ties these together in a coherent whole illustrating this new approach to the social sciences."--Brian Skyrms, University of California, Irvine

"This book is leading what is likely to be an increasingly important line of thought. The central argument and its illustrative applications present conceptual and methodological innovations that clearly have enormous potential. The writing is concise, accurate, balanced, and entertaining. Readers will be broadened, challenged, provoked, and inspired."--John Steinbruner, University of Maryland

See all Editorial Reviews

Product Details

• Hardcover: 352 pages
• Publisher: Princeton University Press (January 8, 2007)
• Language: English
• ISBN-10: 0691125473
• ISBN-13: 978-0691125473
• Product Dimensions: 9.3 x 6.2 x 1.2 inches
• Shipping Weight: 2.2 pounds (View shipping rates and policies)
• Average Customer Review: 4.7 out of 5 stars  See all reviews (6 customer reviews)
• Amazon Best Sellers Rank: #760,709 in Books (See Top 100 in Books)

More About the Author

Discover books, learn about writers, read author blogs, and more.

Customer Reviews

Most Helpful Customer Reviews

29 of 30 people found the following review helpful:

5.0 out of 5 stars A Landmark Publication, March 7, 2007

By 

Andrew Ilachinski (Springfield, VA USA) - See all my reviews
(REAL NAME)   

Amazon Verified Purchase

This review is from: Generative Social Science: Studies in Agent-Based Computational Modeling (Princeton Studies in Complexity) (Hardcover)

Josh Epstein's new Opus is a landmark publication in the emerging field of multiagent-based simulation of dynamic social systems. Since Josh is not only one of this still nascent (though burgeoning) field's ablest and most creative practitioners, but also among its most thoughtful critics, the reader of has two treats in store: (1) a generous, and wide-ranging, sampling of case studies (including social networks and evolution, population growth, emergence of economic classes, civil unrest, timing of retirement, the dynamics of adaptive organizations and the spread of infectious disease), and (2) a cogent "meta" discussion of what multiagent models ARE, ARE NOT and how (when their properties and limitations are *not* properly taken account of) they can easily be MISAPPLIED.

Far from suggesting that multiagent-based models are a panacea solution to all (or most) social dynamical systems, Josh's book carefully articulates the conditions for which such an approach IS (and is NOT) appropriate; an approach rarely taken by other, similar, overviews of the field. Indeed, the cogent philosophical discussion in Chapter One - alone! - in which the generativist's position is defined and put into a broader modeling/simulation context, is worth the price of admission; I have not seen a better "manifesto" of multiagent-based modeling elsewhere.

Finally, without taking away any of the inherent "beauty" (in the technical sense) of the often exaggerated concept of "emergence," Josh succeeds admirably in both defining the term, and de-mystifying it, stripping it of some of its unnecessary "quasi-mystical" baggage (at least as it is often portrayed in lay publications).

Anyone who is interested in understanding how agent models may be used to help explore the dynamics of social dynamical systems, should have this book firmly on top of their "must read" list! Josh has generously provided future generations of agent explorers their go-to source of both inspiration and ideas. Well done Josh!

8 of 8 people found the following review helpful:

5.0 out of 5 stars Excellent example of cross-disciplinary social science using theory, August 7, 2007

By 

Robert J. Muller (San Francisco, CA USA) - See all my reviews
(REAL NAME)   

This review is from: Generative Social Science: Studies in Agent-Based Computational Modeling (Princeton Studies in Complexity) (Hardcover)

It's refreshing and exciting, in a quiet intellectual kind of way, to encounter a book that includes philosophy of science, music theory, Anasazi disappearance mysteries, ethnic cleansing, and an explanation of why CEOs exist. Josh has produced the book I've been wanting to read any time during the last 20 years, which have been a bit barren from the theory and modeling perspective in social science. He also makes clear the mathematical and philosophical basis of the agent-based approach, producing a baseline both for future work in the field and for competing paradigms such as systems dynamics, discrete simulations, and cellular automata (Wolfram's New Kind of Science), however incommensurable. I was particularly interested in the occasional use of probability modeling (negative exponential distributions generated through simple rules are a very interesting advance in understanding the waiting times between civil violence outbursts) and I'd love to see a deeper relationship established, say between Bayesian models of dynamic systems and agent-based models. Keep up the great work, Josh! Also, kudos to the publisher for the sheer quality of the book: excellent paper, great color plates, and priced to sell rather than as the work of art it is.

7 of 7 people found the following review helpful:

5.0 out of 5 stars Excellent survey of the author's work, July 27, 2007

By 

O. Litwin "orendog" (Norwalk, CA United States) - See all my reviews
(REAL NAME)   

This review is from: Generative Social Science: Studies in Agent-Based Computational Modeling (Princeton Studies in Complexity) (Hardcover)

This book did a good job of introducing me to the current state of agent-based modeling. It also, perhaps inadvertently, highlighted some of the current weaknesses of the field. In particular, the models shown in each paper rarely shared common features, and there was little consistency in method.

Epstein argues persuasively that agent-based modeling is a tool, not a methodological approach, and you should no sooner expect consistent usage here than with differential calculus. That said, it was a bit disconcerting.

Also, while the goal espoused here was to use the bare minimum of constraints that retain explanatory power, I was disappointed that relevant work from other fields was often abstracted away. For example, a few models used social networks; but the networks presented were static, not dynamic, and were not built around power-law ratios. Such additional complexity may well have distracted from the main point; but it would have been nice to see at least some discussion of why the models were simplified.

Regardless, I was very pleased with the book and would highly recommend it.