Emergence as the conversion of information: a unifying theory

Area of Science:

• Complex systems science
• Information theory
• Philosophy of science

Background:

• Reductionism is a common scientific strategy, but its universal applicability is questioned by the existence of special sciences.
• Previous work indicated dimensionality reduction can increase causal emergence, suggesting macroscales may alter system dynamics.

Purpose of the Study:

• To develop a mathematical framework for emergence based on information conversion.
• To investigate how coarse-graining affects information types within a system.
• To challenge the notion of universal reduction in scientific modeling.

Main Methods:

• Developed an information conversion framework for emergence.
• Applied mutual information measures to Boolean networks.
• Utilized partial information decomposition to analyze information atoms (redundant, unique, synergistic).
• Introduced a novel measure for synergy bias in information decomposition.

Main Results:

• Demonstrated that coarse-graining can convert information from one type to another.
• Showed that the synergistic information component can increase at macroscales.
• Proved information conversion occurs even when total mutual information remains constant between scales.
• Evidence suggests reduction can lead to a loss of synergistic information.

Conclusions:

• The proposed framework complexifies the idea of universal reduction in science.
• Reductionist strategies may lead to a loss of crucial synergistic information in scientific models.
• Emergent phenomena may arise from information conversion during coarse-graining, not just increased dependency.