Naturalism, tractability and the adaptive toolbox.

Area of Science:

• Cognitive Science
• Decision Making
• Philosophy of Mind

Background:

• Recent research highlights human ability to achieve epistemic success using simple heuristics and minimal information.
• This success is attributed to leveraging environmental features, suggesting a naturalized view of rationality.
• The 'ought-can' principle supports this view, aligning intuitive rationality with practical application.

Purpose of the Study:

• To investigate the computational feasibility of the adaptation process underlying heuristics-based ('ecological') rationality.
• To determine if the naturalization of rationality, based on simple heuristics, is computationally tractable.
• To examine the implications of computational intractability for the project of naturalizing rationality.

Main Methods:

• Analysis of the computational resources required for adaptation processes that guarantee rationality.
• Evaluation of various auxiliary assumptions proposed to mitigate computational burdens.
• Assessment of the robustness of computational intractability in adaptation.

Main Results:

• The adaptation process for ecological rationality demands astronomical computational resources, rendering it intractable.
• Tested auxiliary assumptions failed to resolve the computational intractability.
• Computational intractability was found to be a robust property of adaptation.

Conclusions:

• The naturalization of rationality via simple heuristics is computationally endangered.
• The project of naturalizing rationality faces significant challenges due to the intractability of the underlying adaptation processes.
• Findings question the feasibility of a computationally grounded, naturalized account of human rationality.