Physics and Biology: towards a unified model

摘要

Biological systems have a vertical architecture that allows them to exploit micro-physical dynamics for information processing and related macroscopic functions. Macroscopic sensory information in this vertical picture is transduced to increasingly microscopic forms within biological cells and then back to macroscopic form. Processing of information can occur at any level of organization, but much of the most powerful processing occurs at the molecular and submolecular level. The open process of Darwinian evolution plays an important role, since this provides the mechanism for harnessing physical dynamics for coherent function. The vertical architecture is analogous to a quantum measurement system, with transduction of input signals corresponding to state preparation and amplification of the microstate corresponding to measurement. The key point is that the microphysical dynamics is not classically picturable, whereas the macroscopic actions are definite and picturable. If this analogy is taken seriously, it becomes necessary to suppose that irreversible projection processes occur in organisms, despite the fact that the standard equations of motion are reversible. We construct a model that embeds such irreversible measurement interactions into interactions that mediate the conventional forces. The idea is that the forces between observable particles depend on the density of negative energy particles in a surrounding Dirac type vacuum. In systems that are not overly macroscopic or overly far from equilibrium this dependence is hidden, and as a consequence the force appears conservative. The model suggests that the irreversible aspect of the dynamics played an important role in the early universe, but became masked in ordinary laboratory situations as the structure of the vacuum and the distribution of mass and charge equilibrated. Organisms are particularly effective at unmasking the underlying irreversibility due to their sensitive amplification mechanisms. Unifying measurement and force type interactions makes it possible for physical models to fit more naturally to models of cognition.