Fitness Over Truth / Evolutionary Biology

What selection actually optimizes

Selection has exactly one currency: surviving copies. A trait that helps an organism reproduce spreads, whether or not it tracks reality. A trait that tracks reality perfectly but costs a reproductive edge does not.

So the question “did evolution build us to see the truth?” has a precise answer: only where seeing the truth happened to pay, and only as much as it paid. Truth is not the target. It is, at best, an occasional instrument that selection picks up when useful and drops when expensive.

Perception as interface, not window

It is tempting to think perception is a window — that evolution refined our senses toward an ever more faithful picture of the world. But fidelity is costly, and the organism does not need fidelity. It needs to act.

Think of a desktop interface. The icon for a file is not what the file is — not the voltages and magnetic states on the disk. It is a useful fiction that hides the truth so you can act without drowning in it. Perception is like this. Colour, solidity, the bounded “objects” we see — these are species-specific shorthand tuned to guide behaviour, not transcriptions of what is there. A creature that saw reality “as it is,” in full, would be slower than a creature that saw a convenient cartoon and ran. The cartoon won.

Bias is not malfunction

This reframes the long catalogue of cognitive biases. We treat them as bugs — failures of an instrument meant to be rational. Mostly they are not bugs. They are the instrument working as selected.

We overweight vivid risks and underweight statistical ones, because the ancestor who fled the rustle that might be a predator out-reproduced the one who waited for the data. We see agency in noise, because a false alarm is cheap and a missed leopard is fatal. We are wired to believe what our group believes, because exile was a death sentence and truth was rarely worth it. Each of these is a good bet under ancestral conditions. Each is also, reliably, a way of being wrong.

The mind is not a broken truth-detector. It is a well-tuned fitness-detector that we are now asking, against its design, to detect truth.

The gap we live in

Where fitness and truth coincided, our intuitions are excellent — catching a thrown object, reading a face, judging a slope. Where they diverged, our intuitions are confidently, systematically wrong — probability, large numbers, deep time, the behaviour of systems with no ancestral analogue.

Modernity is mostly built out of exactly the second category. Statistics, compound interest, exponential growth, abstract institutions, the long-term consequences of collective action — none of these were present in the environment that shaped our default cognition. The factory settings are out of distribution.

Why methods exist

This is the quiet justification for every disciplined way of knowing humans have ever invented. Logic, mathematics, controlled experiment, double-blind trials, peer review, written record, the slow norms of science — these are prosthetics for a mind that was not built to be right. They are external scaffolds that catch the errors our wiring reliably produces.

To trust your unaided intuition is to trust an instrument calibrated for a world that no longer exists, toward a goal — survival — that is not the goal you now have. Knowing this does not free you from the biases; you cannot feel your way out of a wired-in default. It only tells you where to distrust yourself, and to reach for the scaffold.

The same logic that explains why we misperceive the world also explains why we misperceive our own wants — and who, exactly, those wants are working for.