Systems thinking books for people who study societies

If you read one systems thinking book, make it Donella Meadows's Thinking in Systems, a short primer on stocks, flows and feedback loops that suits anyone, with or without math. For the wider science of complex systems, read Melanie Mitchell's Complexity: A Guided Tour next; it suits readers who want the ideas explained properly and can take a chapter on computation. For the step from systems to societies, the classic is Thomas Schelling's Micromotives and Macrobehavior: essays on how sensible individual choices add up to patterns nobody chose, including the checkerboard segregation model that modeling textbooks still teach.

The books were chosen from the publishers' descriptions and tables of contents and from reviews in Physics Today, Cosmos + Taxis and the Journal of Artificial Societies and Social Simulation (JASSS). The angle is that of someone who studies societies, not fisheries and factories.

Pick by what you want

What a systems book can do for an anthropologist

Systems thinking, in Meadows's sense, sees stocks, flows and loops. A stock is anything that accumulates, such as grain in a store; flows fill and drain it; a feedback loop is a stock changing its own flows, either feeding on itself (reinforcing) or pushing back toward a limit (balancing). Complexity science asks the same question from the other end: how do many agents following local rules produce order nobody planned?

Both ideas run through cultural evolution. Conformist transmission, copying whatever most people around you do, is a reinforcing loop, worked out in our explainer on transmission biases. A simpler hypothetical has the same shape. In a valley of 100 households, 20 have taken up a new way of storing seed. Each season, every household that has not adopted it switches with a chance equal to the share that already has. Rounding to whole households:

New adopters climb while the reinforcing loop is in charge, then fall as the holdouts run out and a balancing loop takes over. That S-curve is the shape Page's chapters on diffusion work with, and it turns up in many simulations of innovation. Meadows and West mostly work with totals like this; the classics and textbooks below work with individuals and who sits next to whom. Anthropologists usually need both.

Start here

Thinking in Systems, Donella H. Meadows (2008)

Meadows was the lead author of The Limits to Growth (1972). She drafted this primer in 1993 and died in 2001; colleagues at the Sustainability Institute, with Diana Wright as editor, shaped the draft into the book Chelsea Green published in 2008. It moves from system structure and a "systems zoo" of small examples, through why systems work, surprise us and fall into traps, to leverage points, where a small push changes a lot. It suits anyone. The caveat here: its examples lean toward environment and economics, and people appear as quantities in a stock. Kin, neighbors and culture are left to you.

Complexity: The Emerging Science at the Edge of Order and Chaos, M. Mitchell Waldrop (1992)

This is narrative journalism about the founding years of the Santa Fe Institute, where scientists from many fields looked for common rules of complex adaptive systems. Waldrop carries the ideas through people, among them the economist Brian Arthur, the biologist Stuart Kauffman and Chris Langton of artificial life. It suits readers who like history. Caveats: it records a moment the science has moved past, and a Publishers Weekly review at the time felt the attention to personalities weakened the explanations.

The Model Thinker, Scott E. Page (2018)

Page teaches complex systems, political science and economics at the University of Michigan. His case is for many-model thinking: put several simple models to work on one problem and look at where they agree. Most of the 29 chapters take one family of models, among them power-law distributions, networks, diffusion and contagion, local interaction, Markov models, system dynamics, feedback and cooperation. It suits social scientists who want many models in one volume. Caveats, from Ted Lewis's review in Cosmos + Taxis: the opening chapters run long, the network chapters feel dated, and there is no index.

Going further

Complexity: A Guided Tour, Melanie Mitchell (2009)

Mitchell, a computer scientist, wrote this for general readers; it won the 2010 Phi Beta Kappa Book Award in Science. The tour takes in dynamics, chaos, information, computation and evolution; genetic algorithms, her own research field; cellular automata and Axelrod's work on cooperation; and network thinking, including the small-world idea in our explainer on six degrees. One chapter asks how complexity might be measured and finds no single agreed answer. Caveat: John Bragin's review in JASSS counted the lack of a chapter on agent-based simulation as a real gap, so simulation-minded readers will want a textbook alongside.

Scale, Geoffrey West (2017)

West, a theoretical physicist who was president of the Santa Fe Institute from 2005 to 2009, starts from biology: an animal's metabolic rate grows roughly as the 3/4 power of its mass, so an animal 16 times heavier needs about 8 times the energy (16^(3/4) = 8). In Jayanth Banavar's summary for Physics Today, measures such as economic output, patents and crime grow with city population at an exponent of about 1.15, so doubling a city gives roughly 2^1.15 ≈ 2.2 times as much of each, good and bad. Archaeologists have taken up the idea: Ortman and colleagues (2014) found settled area across more than 1,500 settlements in the pre-Hispanic Basin of Mexico growing more slowly than population, as urban scaling theory predicts. Caveats: Banavar argues that West presents his fractal-network explanation of the 3/4 law as settled when others dispute it, and the book runs close to 500 pages.

Anthropologists meet power laws in humbler places too. Hahn and Bentley (2003) found that American baby names follow one in every decade of the twentieth century, and that random copying reproduces it: the neutral model of cultural drift, a good null hypothesis before any grander explanation.

Classics and model builders

Micromotives and Macrobehavior, Thomas C. Schelling (1978; updated edition 2006)

Schelling, who shared the 2005 Nobel Prize in economics, wrote these essays on one theme: reasonable individual choices that sum to something nobody wanted. The best-known case is his segregation model, published in 1971. Households of two kinds sit on a grid, and each moves if too few neighbors are like itself. Even when everyone would accept a mixed street, the grid can end up almost completely sorted. The updated Norton edition adds a new preface and Schelling's Nobel lecture. Caveats: the examples belong to their decade, and the model shows that mild preferences are enough to produce segregation, not that they caused it in any real town.

An anthropological footnote: Rainer Hegselmann's history of the model (2017) shows that James Minoru Sakoda published a more general checkerboard model in the first issue of the same journal in 1971, with a precursor in his 1949 Berkeley dissertation. Sakoda based that dissertation on his work as a participant observer for a Berkeley study of the wartime camps for Japanese Americans, in which he was himself held, at Tule Lake and later Minidoka. One of the earliest models of its kind grew out of fieldwork done under confinement.

Growing Artificial Societies, Joshua M. Epstein and Robert L. Axtell (1996)

The founding book of Sugarscape. Agents with a vision range, a metabolism and a store of sugar live on a grid, and the core rule is to look as far as you can see, move to the richest spot and eat. Add seasons and they migrate. Add a second resource, spice, with trade, and something like a market appears, along with inequality. On a map with two sugar peaks, culturally distinct groups form around each and later compete. It suits readers who want the generative argument at full strength: rules that can grow a pattern are one candidate explanation for it. Caveat: its agents live nowhere in particular. For the archaeological sequel in Long House Valley, see our explainer on agent-based models in anthropology.

Textbooks for building your own

All three use NetLogo, the free modeling environment written by Uri Wilensky.

Uri Wilensky and William Rand, An Introduction to Agent-Based Modeling: Modeling Natural, Social, and Engineered Complex Systems with NetLogo (MIT Press, 2015), teaches the language alongside the theory, up to verification and validation, for advanced undergraduates. Jennifer Badham's JASSS review warns that some exercises use techniques not yet taught.

Steven F. Railsback and Volker Grimm, Agent-Based and Individual-Based Modeling: A Practical Introduction (Princeton University Press, second edition 2019), puts more weight on designing a model around a question and testing it against observed patterns (pattern-oriented modeling). Both authors are ecologists, and building from observation will suit fieldworkers.

Paul E. Smaldino, Modeling Social Behavior: Mathematical and Agent-Based Models of Social Dynamics and Cultural Evolution (Princeton University Press, 2023), sits closest to this site: segregation, contagion, opinion dynamics, cooperation, norms and networks, each taught through a simple model, with equations and agents side by side.

A reading order

Meadows for loops, Schelling for the jump from motives to patterns, Mitchell for the science around both, then Smaldino or Wilensky and Rand. Keep Waldrop and West for evenings and Page for reference. The reading shelf gathers every book the site recommends. A first exercise once NetLogo is installed: rebuild the seed-storage table with 100 agents, run it twenty times, and count how many runs match the table season by season.

^† Links to shops are affiliate links: if you buy through one, we may earn a commission.

Sources and further reading

• Banavar, Jayanth R. (2018). "An intriguing take on complexity" (review of West, Scale). Physics Today 71 (1): 57. PDF
• Badham, Jennifer (2015). Review of Wilensky and Rand, An Introduction to Agent-Based Modeling. Journal of Artificial Societies and Social Simulation 18 (4). jasss.org
• Bragin, John (2010). Review of Mitchell, Complexity: A Guided Tour. Journal of Artificial Societies and Social Simulation 13 (2). jasss.org
• Epstein, Joshua M., and Robert L. Axtell (1996). Growing Artificial Societies: Social Science from the Bottom Up. Brookings Institution Press and MIT Press. Publisher's page
• Hahn, Matthew W., and R. Alexander Bentley (2003). "Drift as a mechanism for cultural change: an example from baby names." Proceedings of the Royal Society B 270 (Suppl. 1): S120-S123. PMC
• Hegselmann, Rainer (2017). "Thomas C. Schelling and James M. Sakoda: The Intellectual, Technical, and Social History of a Model." Journal of Artificial Societies and Social Simulation 20 (3): 15. jasss.org
• Lewis, Ted G. (2019). Review of Page, The Model Thinker. Cosmos + Taxis 6 (6+7). PDF
• Ortman, Scott G., Andrew H. F. Cabaniss, Jennie O. Sturm, and Luís M. A. Bettencourt (2014). "The Pre-History of Urban Scaling." PLOS ONE 9 (2): e87902. journals.plos.org