Physical causality and brain theories.

The history of deterministic theories in physics is reviewed, and four levels of determinism are found: 1) absolute, 2) asymptotic, 3) probabilistic, and 4) absolute indeterminism. Nagel's view that all causal laws are deterministic in the frame of the state descriptions to which they refer is acknowledged, but the inevitability of macroscopic measurement noise may hint that dynamical laws are innately noisy. Quantum mechanical effects are not the noise source. Symmetry and broken symmetry are introduced as physical concepts that can account both for lawfulness, and for the hierarchical nature of the universe. Physical ideas are chosen over those of formal systems with indirect self-reference as the basis of a global theory of brains. By exclusion it is concluded that only a statistical thermodynamics, combined with nonlinear mechanics, has the features needed for theorizing about brains in a physical sense. Quantum mechanics is judged not to be relevant. New statistical thermodynamic theories are briefly described, and their strengths and weaknesses noted. The question, "Why should neuroscience look to physics for its theories?" is raised and answered. Some concrete objectives for a program of theoretical research are stated.