Original title: Evolutionary Theory Makes Predictions About Cognition Beyond Rational Optimization
Cameron Turner, Thomas L. Griffiths · September 2026
Scientists often assume that evolution designs animal minds to make perfectly rational choices, like gathering accurate information or maximizing expected rewards. However, the researchers argue that evolution does not care about individual accuracy. Instead, evolution favors whatever keeps an entire family line alive over many generations. To test this, the researchers built a signal detection model based on lineage survival. They found that in constantly changing environments, minds actually evolve weak prior beliefs. This makes individuals make less accurate decisions, but it causes offspring to act differently from one another. That behavioral diversity guarantees at least some offspring survive new challenges. Natural selection shapes minds around stable trends, but changing environments strip cognitive structure away to generate variety. Ultimately, evolution maximizes the geometric mean of offspring over generations.
Imagine packing a suitcase for a trip to an unpredictable climate. If you pack exactly according to the weather forecast, you might be comfortable on some days but unprepared if conditions shift. A better strategy is to pack a variety of items, even if some seem unnecessary, so you have something suitable no matter what. Here, evolution's objective of maximizing offspring over generations corresponds to this variety strategy: instead of optimizing for one environment, it shapes the mind to produce diverse behaviors, ensuring some offspring can handle novel challenges.
This work challenges the assumption that minds are optimized for accuracy. It suggests evolution may shape cognition to prioritize variation in changing environments, explaining why humans sometimes make seemingly irrational decisions. That insight could improve AI and our understanding of cognitive biases.
The abstract does not provide experimental validation or specific predictions for real organisms. The model is limited to signal detection and may not generalize to all cognitive tasks. It also does not specify the exact mathematical conditions under which weak priors are favored beyond change-dominated environments.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Evolutionary Theory Makes Predictions About Cognition Beyond Rational Optimization — Cameron Turner, Thomas L. Griffiths — PsyArXiv — doi:10.31234/osf.io/vjrkc_v2
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.