The evolutionary origin of Bayesian heuristics and finite memory

Area of Science:

• Cognitive Science
• Evolutionary Biology
• Computational Neuroscience

Background:

• Bayes' rule is a foundational principle across disciplines, but its cognitive origins and generalization basis remain unclear.
• Existing research often assumes rational agents, overlooking evolutionary explanations for cognitive strategies.

Purpose of the Study:

• To investigate the evolutionary emergence of Bayesian inference as a cognitive strategy.
• To identify environmental conditions favoring the development of finite memory in decision-making.

Main Methods:

• Utilized a simple binary choice model subjected to natural selection.
• Derived Bayesian inference as an adaptive behavior in specific stochastic environments.
• Analyzed environments favoring finite memory: Markov, nonstationary, and informationally limited.

Main Results:

• Demonstrated that Bayesian inference can evolve in populations of non-reasoning individuals.
• Identified specific environmental characteristics (Markov, nonstationary, information-limited) that promote finite memory.
• Showed that evolved behavior can deviate from optimal Bayesian strategies.

Conclusions:

• Evolutionary processes can establish Bayesian inference as an adaptive cognitive strategy.
• Finite memory emerges naturally under certain environmental conditions, explaining cognitive limitations.
• Provides an evolutionary framework for understanding human cognitive phenomena and deviations from ideal Bayesian processing.