Ternary Logic as the Natural Substrate for Biological Computation: Information-Theoretic Foundations and the Resolution of 35-Year Model-Based RL Instability

Binary computation enforces a false dichotomy on naturally ternary systems. We derive information-theoretic proofs establishing that balanced ternary encoding (-1, 0, +1) achieves a log₂(3)/log₂(2) = 1.585 bit-per-symbol advantage, representing 58.5% greater information density than binary. We further demonstrate that model-based reinforcement learning — which has resisted practical implementation for 35 years in binary architectures — achieves stable 63-step planning horizons when implemented on ternary substrates, compared with the 3–5 step collapse characteristic of binary systems.