Selection
Most people explain complex things by pointing to a designer. The eye is intricate, so someone must have designed it. Markets are efficient, so someone must be coordinating them. Language is structured, so someone must have planned it.
Darwin’s insight was simpler and stranger: you don’t need a designer. You need three ingredients operating over time — variation, selection pressure, and inheritance. Given these, complex functional order assembles itself.
The algorithm is almost embarrassingly simple:
- Variation — Copies differ from originals. Mutations, errors, recombinations, experiments.
- Selection — Some variants survive and reproduce better than others. The environment filters.
- Inheritance — Survivors pass traits to the next round. What works accumulates.
Repeat for a million generations. What emerges looks designed — exquisitely fitted to its environment — but no one designed it. The giraffe’s neck, the octopus’s eye, the bacterial flagellum. All built by the same blind process: try things, keep what works, discard the rest.
The key insight: selection doesn’t optimize for anything in particular. It optimizes for survival in this environment. Change the environment, and what gets selected changes.
Biology is the obvious case, but selection operates wherever the three ingredients exist.
Ideas: Variations emerge in thought → some spread through social adoption → successful memes get copied and modified. Richard Dawkins named this, but the pattern predates the word. Religions, scientific theories, political ideologies — all subject to memetic selection. What survives isn’t necessarily true. It’s what spreads.
Markets: Entrepreneurs try variations → customers choose → surviving businesses get imitated. The market “discovers” prices not through calculation but through selection. Hayek called this a “discovery procedure” — no one knows enough to plan an economy, but selection finds viable arrangements anyway.
Technology: Tinkerers try modifications → use reveals what works → improvements get retained. Brian Arthur traces how technologies evolve through accumulated small changes, each selected by usefulness. The bicycle took decades of variation and selection to reach its current form.
Learning: The brain generates action possibilities → outcomes provide feedback → successful patterns strengthen. Gerald Edelman called this “neural Darwinism” — the brain learns through accumulated trial and feedback rather than explicit instruction.
Culture: Practices vary across communities → some communities thrive → successful practices spread. This is how traditions accumulate wisdom without anyone understanding why they work. Chesterton’s fence contains selection’s legacy.
Selection explains what design cannot: functional complexity without intent.
Watch a skilled potter. Their technique looks designed — each movement precise, the whole coordinated. But they didn’t design it. Thousands of hours of variation (trying movements), selection (feeling what works), and inheritance (the body remembers) produced this. The same process that shaped the potter’s hands over evolutionary time shaped their skill over developmental time.
This reframes many puzzles:
Why do institutions resist reform? Because they’ve been selected for survival in their environment, not for their stated purpose. The bureaucracy that seems dysfunctional may be exquisitely adapted to internal political pressures you don’t see.
Why do good ideas sometimes lose? Selection pressure isn’t always aligned with truth or quality. Ideas spread based on what makes them spread — emotional resonance, simplicity, tribal signaling — not on correctness.
Why do traditions contain wisdom we can’t articulate? They survived long selection processes. The practice that seems arbitrary may encode solutions to problems we’ve forgotten we had. Lindy effects are selection effects.
Why do markets work despite no one understanding them? Because markets are selection processes, not designed systems. They don’t need someone to understand them. They need variation (entrepreneurship), selection (consumer choice), and inheritance (imitation of success).
Selection has limits and dark sides.
Local optima: Selection finds peaks, but not necessarily the highest peak. It can get stuck on local maxima, unable to cross valleys of lower fitness to reach higher ground. This is why radical innovation often comes from outsiders — they start from different positions.
Lag: Selection adapts to past environments. When conditions change faster than selection can respond, you get maladaptation. Evolutionary mismatch: bodies adapted to scarcity now live in abundance. Institutional mismatch: organizations adapted to last decade’s problems.
Selection of the selector: What does the selecting? In markets, consumer preferences — but those preferences are themselves shaped by selection. In culture, attention — but attention is captured by what spreads. The meta-level matters.
Path dependence: Selection doesn’t start fresh. It builds on what exists. This creates lock-in — the QWERTY keyboard persists not because it’s optimal but because switching costs exceed benefits. History constrains possibility.
The deepest implication: most of what exists was never designed.
Not just species. Institutions, traditions, languages, skills, markets, cities, ecosystems. All shaped primarily by selection, not by intention. The designed parts float on a vast sea of evolved order.
This suggests humility. When you encounter something complex that works, don’t assume you understand why it works. It may have been selected for reasons invisible to you. And when you try to redesign it, you’re not just fighting inertia — you’re overriding accumulated solutions with your limited understanding.
High modernism failed not because planners were stupid, but because they replaced evolved order with designed order. The designed version had fewer flaws the planner could see, but also fewer solutions the planner hadn’t thought of.
Selection is slow, wasteful, and blind. It’s also the only process that reliably builds complex functional order from scratch. Respect it.
Go Deeper
Books
- The Blind Watchmaker by Richard Dawkins — The clearest explanation of how selection produces apparent design. The extended phenotype chapter alone is worth the book.
- The Origin of Wealth by Eric Beinhocker — Selection applied to economics. How markets work as evolutionary systems.
- The Evolution of Everything by Matt Ridley — Selection beyond biology. Culture, morality, technology, economy — all evolving.
- Antifragile by Nassim Nicholas Taleb — What survives selection gains from volatility. The link between selection and robustness.
- Darwin’s Dangerous Idea by Daniel Dennett — Selection as “universal acid” that transforms every field it touches. Philosophically ambitious.
Related: [[emergence]], [[adjacent-possible]], [[lindy-effect]], [[antifragility]], [[chestertons-fence]], [[bricolage]], [[scenius]], [[ecological-succession]], [[tacit-knowledge]]