Growing Artificial Societies: Social Science from the Bottom Up

by Joshua M. Epstein, Robert L. Axtell

How do social structures and group behaviors arise from the interaction of individuals? In this groundbreaking study, Joshua M. Epstein and Robert L. Axtell approach this age-old question with cutting-edge computer simulation techniques. Such fundamental collective behaviors as group formation, cultural transmission, combat, and trade are seen to "emerge" from the interaction of individual agents following simple local rules.

In their computer model, Epstein and Axtell begin the development of a "bottom up" social science. Their program, named Sugarscape, simulates the behavior of artificial people (agents) located on a landscape of a generalized resource (sugar). Agents are born onto the Sugarscape with a vision, a metabolism, a speed, and other genetic attributes. Their movement is governed by a simple local rule: "look around as far as you can; find the spot with the most sugar; go there and eat the sugar." Every time an agent moves, it burns sugar at an amount equal to its metabolic rate. Agents die if and when they burn up all their sugar. A remarkable range of social phenomena emerge. For example, when seasons are introduced, migration and hibernation can be observed. Agents are accumulating sugar at all times, so there is always a distribution of wealth.

Next, Epstein and Axtell attempt to grow a "proto-history" of civilization. It starts with agents scattered about a twin-peaked landscape; over time, there is self-organization into spatially segregated and culturally distinct "tribes" centered on the peaks of the Sugarscape. Population growth forces each tribe to disperse into the sugar lowlands between the mountains. There, the two tribes interact, engaging in combat and competing for cultural dominance, to produce complex social histories with violent expansionist phases, peaceful periods, and so on. The proto-history combines a number of ingredients, each of which generates insights of its own. One of these ingredients is sexual reproduction. In some runs, the population becomes thin, birth rates fall, and the population can crash. Alternatively, the agents may over-populate their environment, driving it into ecological collapse.

When Epstein and Axtell introduce a second resource (spice) to the Sugarscape and allow the agents to trade, an economic market emerges. The introduction of pollution resulting from resource-mining permits the study of economic markets in the presence of environmental factors.

This study is part of the 2050 Project, a joint venture of the Santa Fe Institute, the World Resources Institute, and the Brookings Institution. The project is an international effort to identify conditions for a sustainable global system in the middle of the next century and to design policy actions to help achieve such a system.

Genres: Sociology, Nonfiction, Science, Social Science, Economics, Technology, Computers

208 pages, Paperback

First published November 1, 1994

Reviews

[Brian Powell · Rating 3 out of 5]

I hate to be tough on this book, because it really is a pretty decent work. The authors demonstrate, through gradually increasingly complex scenarios, how high-level phenomena can emerge from a network of heterogeneous agents interacting according to local rules, including those governing trade, migration, war, and disease transmission. Some of these results are obvious and relatively unenlightening; others, as the local rules become more complex, hint at the unexpected.

Ultimately, there appears to be a payoff to using bottom-up methodologies to understand and explain complex societal phenomena, but there's no free lunch. Indeed, while top-down approaches (where one begins with a phenomenon to be explained and attempts to model it) are intractable in all but the most idealized of scenarios, this difficulty isn't really dissipated by moving to bottom-up models. Rather, it is transformed into thorny questions about rules, properties of agents, interaction dynamics, initial conditions, and the structure of the environment. The authors demonstrate that you can get started down this road by establishing relatively simple settings for these things (that you can sort of get a free lunch), but the resulting phenomena tend to be dry and uninspired, of the kind perhaps manageable in a top-down approach.

Furthermore, the two "opposing" methodologies (top-down/bottom-up) aren't really. With the bottom-up approach espoused in the book, you don't have any idea what you're gonna get. You add ingredients and just start stirring: the stew at the end could taste like god-know-what. And, so, on a practical level, if you're interested in understanding how, say, divorce rates interact with gun violence or poverty or are affected by social safety nets, you'd have to get really lucky with the bottom-up approach to generate precisely these social phenomena from your model. This doesn't make these models useless by any means, but it does seem to suggest that the selling point of these models--that they are fantastically more manageable than difficult top-down approaches--is a tad disingenuous.

In all, though, the book is surely of interest to those curious about local, agent-based computer simulations of societies. The authors demonstrate the plausibility of such approaches as a tool for understanding what goes into complex social phenomena. But it's only plausibility: the reader is left wanting for a more decisive test-drive of a perhaps more mature discipline.

[Maged Essam · Rating 4 out of 5]

I used it as a reference for Agent-based computational economics course. It is really good. It introduces Sugarscape model, its main features, and experiments done through it. Even though it is simple, too simple that it puts reality aside, but I enjoyed the methodology

[Max Bittker · Rating 2 out of 5]

I'm reading this in the wrong decade! love the concepts and derived work since.

[Josephene · Rating 4 out of 5]

Read this book because of Making Sense of Chaos by J. Doyne Farmer. Growing Artificial Societies is a solid introduction into the basics of agent-based modeling, what it’s capable of, and what insights you can gain from it. As others have mentioned, the content is dated at this point, but it’s gotten me interested in reading more recent work in this area. The book is technical but I kinda wished I could just read some code or software documentation. I did enjoy the structure of the book and how it introduced new ideas and concepts incrementally, like a story, and I liked how it brought more humanity and collectiveness into economics in particular. I wonder how the development of AI and LLMs is impacting generative social science, if any additional insight can be gained from coupling them with agent-based modeling.

[Alexander · Rating 4 out of 5]

A bit dated at this point, but the underlying ideas are really engaging and intriguing. The writing was generally exciting enough to keep me reading, and their results (even from relatively simple models) raise some great questions and ideas.

[Aarav · Rating 4 out of 5]

Ever wanted to play god, but like, scientifically? Epstein and Axtell basically hand you the keys to a tiny, simulated civilization and say, Go on, then. Watch your little digital people form economies, start wars, and struggle to survive, all while you take notes like an omniscient anthropologist. This book is all about agent-based modeling, which sounds intimidating but is actually just the process of making a bunch of simple rules and seeing what happens when you let them loose. And what happens is fascinating. Inequality? Emergent. Social norms? Self-reinforcing. Entire civilizations collapsing because of bad resource management? Tragically relatable. It’s both a technical and philosophical rabbit hole.