Information Theory of Biological Integration
Henry Quastler · The American Naturalist · 1959
Two phenomena are commonly observed in complex systems: integration, or the restriction of possible configurations of the system through strong coherence between its parts; and feature sampling, or the restriction of functional significance to a fraction of the aspects (the "signature") of a complex configuration. The two effects complement each other: integration means over-determination of systems configurations, and feature sampling, under-evaluation. The stability of a system depends on degree and kind of both effects. Assuming the specification of admissible configurations, and of signatures, to be random, the degree of both effects alone determines the stability. This hypothesis leads to a mathematical relation between probability of systems failure and dose of some injurious agent. The relation found fits a wide variety of experimental data.