The Complex World: An Introduction to the Foundations of Complexity Science

An excellent intro to the field and the development of complexity science. It’s one thing to be loosely familiar with complexity concepts such as emergence, feedback, and adaptation, and quite another to understand where they came from and why they matter. Understanding the history of a field helps you to be more critical, to recognise both strengths and limits.

I like how Krakauer frames complexity through a distinction between two world systems. The first world consists of closed, predictable domains governed by symmetry, reversibility, and stable laws, the kinds of systems traditionally studied in classical physics, much of chemistry, and parts of engineering where control, optimisation, and prediction are viable. The second world consists of open, far from equilibrium systems shaped by history, adaptation and emergence. Complexity science (and many organisational challenges) sit squarely in this latter world, often at the uneasy boundary between the two, where attempts to apply tools from the first begin to break down.

The book traces the foundations of complexity across four pillars: thermodynamics, evolution, nonlinear dynamics, and computation. These ideas explain why prediction, equilibrium and optimisation so often fail in organisations and societies. It makes clear why attention must shift away from control and towards feedback, learning, and adaptive capacity.

Krakauer clears up a common misunderstanding that complexity theory is anti-reduction. The issue is a mixing up two meanings of reduction. One is the idea that everything should be explained in terms of basic particles or universal laws. The other is the practical act of simplifying a model to the smallest set of assumptions that still fits the evidence. Complexity science rejects the first but depends on the second. It does not reduce organisations or societies to physics, but it does demand clear, disciplined explanations. Without this distinction, complexity can slide into vague holism, while overcommitment to fundamental explanation strips away what matters at the level where action happens.

While the Santa Fe Institute is best known as the home of computational complexity and agent based modelling, the book consistently argues for pluralism. Emergence operates at multiple levels, so no single theory or method is sufficient. Computational models are powerful, but they have limits, especially in human systems where meaning, narrative, identity and power shape outcomes as much as structure or incentives. This is a key insight that drives our approach at House of Complexity: without integration of social science, philosophy, lived experience, complexity as an applied practices becomes disciplinary and extremely limited. It must be transdisciplinary, going across and beyond any single field.

Complexity science itself did not arise as a single method or discipline. It emerged through a gradual fusion of ideas from physics, biology, engineering, computation and the adaptive sciences. Because the phenomena it studies move fluidly across levels and domains, explanation must do the same. This is not a weakness but a reflection of reality, where order and function arise through self organising processes shaped by history rather than universal laws of optimisation.

Drawing on Wittgenstein, Dilthey, and Kuhn, the book explains why different problems require different ways of thinking, why understanding moves back and forth between parts and wholes, and why paradigms form coherent but sometimes incompatible frames. You cannot use the rules of chess to play Go, nor particle physics to understand organisations. Each paradigm brings its own metaphors, models and standards of explanation.

Complexity science does not replace existing frameworks with a single master lens. It legitimises working across perspectives, choosing the right level of description for the problem at hand, and resisting the false comfort of one size fits all solutions. In a connected and volatile world, pluralism is not optional. It is what effective practice demands.