Instabilities and Information in Biological Self-Organization

Abstract

Two classes of self-organizing systems have received much attention, the statistically unstable systems that spontaneously generate new dynamical modes and information-dependent systems in which nonstatistical constraints harness the dynamics. Theories of statistically unstable systems are described in the language of physics and physical chemistry, and they depend strongly on the fundamental laws of nature and only weakly on the initial conditions. By contrast, the information-dependent systems are described largely by special initial conditions and constraints, and they depend only weakly, if at all, on the fundamental laws. This results in statistically unstable theories being described by rate-dependent equations, while the information-dependent systems are described by rate-independent (nonintegrable) constraints. It is argued that an adequate theory of biological self-organization requires that these two complementary modes of description be functionally related, since the key process in morphogenesis is the harnessing of cellular dynamics by the informational constraints of the gene. This could arise if the triggering role of fluctuations could be displaced by informational constraints in the control of the dynamical behavior. However, the spontaneous replacement of chance fluctuations by deterministic informational codes is itself a serious problem of self-organization. At present the only approach requires complementary modes of description for the molecular informational constraints and for the macroscopic dynamical behavior that they harness.