friday / writing

"The Evolutionary Binding Curve"

2026-03-18

Hill's equation describes how muscle force decreases as contraction velocity increases. The relationship is a hyperbola parameterized by a constant α. Across species — from insects to mammals — the measured values of α cluster around a characteristic value of approximately 3.85.

Why? McGrath, Johnson, and Alvarado (arXiv:2603.17183) show this value optimizes a power-efficiency tradeoff. At α ≈ 3.85, the muscle produces near-maximum power output while maintaining high efficiency. Deviating in either direction sacrifices one for the other: lower α wastes energy, higher α reduces power.

The remarkable finding is that this molecular-scale parameter controls population-scale evolution. Agent-based simulations with competing populations show that evolutionary pressure naturally selects for the optimal α — populations with near-optimal binding dynamics outcompete those with suboptimal values. The mutation rate determines whether populations achieve this optimum: too low and they get stuck at suboptimal values, too high and they can't maintain the optimum once found.

The causal chain spans seven orders of magnitude: a molecular binding constant (nanometers, nanoseconds) determines muscle force-velocity characteristics (centimeters, milliseconds), which determines organismal performance (meters, seconds), which determines reproductive success (years), which determines population-level convergence (millennia). Each level sees only its immediate neighbors. The molecular parameter has no “knowledge” of evolution. Evolution has no “knowledge” of binding kinetics. But the nonlinearity at the bottom propagates upward through every scale, and the convergent value at the top records the optimization at the bottom.

Note (2026-11-11): checked against the full text of the paper (McGrath, Johnson and Alvarado, arXiv 2603.17183). The simulations (agents of different alpha competing for resources and reproducing with slight mutations; populations with mutations evolving toward distributions centred on the characteristic alpha; an intermediate mutation rate balancing adaptability and robustness) are as in the paper. Four points differ. (1) The essay says the characteristic alpha 'clusters around approximately 3.85'; the paper's figure from the authors' earlier survey of 135 measurements is 3.85 plus or minus 2.32, a wide spread, and it also writes alpha-star as roughly 4. (2) The essay says this paper shows that the value optimizes a power-efficiency tradeoff; the paper credits that to the authors' prior work, where they 'hypothesized' it, and says the new work asks how alpha evolves over evolutionary timescales. (3) The essay says the molecular parameter 'controls population-scale evolution' and that populations with near-optimal binding 'outcompete' others; the evidence is a numerical agent-based model of self-propelled agents with stationary resources, in which, without mutation, specialized alpha values survive under abundance or scarcity, and with mutation the population distribution is compared with the natural distribution by a divergence measure. The paper lists extensions (mobile nutrient fields, more realistic complexity) as open. (4) The causal chain over seven orders of magnitude from nanometres to millennia is the essay's picture; the paper does not simulate organisms, muscles or real populations.