What Naturalness Measures: Fine-Tuning and Informational Invariants in Cosmology and Dark Matter

Abstract

Naturalness is commonly presented as an objective constraint on physical theories: a model requiring fine-tuning is judged implausible. This presentation conflates a representation-dependent quantity with an invariant one. A fine-tuning verdict depends on the choice of fundamental parameters, the prior, and the measure convention, so it does not by itself fix a feature of the world. Here, I argue that what is objective is structural: the universality class of the map from parameters to observables, invariant under admissible changes of parametrization and measure convention, and independent of any prior over parameter space; it constitutes an informational invariant. On this account naturalness is neither an aesthetic preference nor an objective probability, but a statement about the distinguishability geometry of the representations through which physics encodes observation. I trace the certainty of naturalness verdicts to a tradition, from Ockham through Dirac and Weinberg, in which parsimony and beauty are taken as guides to truth; modern naturalness inherits that tradition's authority without its successive justifications. The argument is developed in the gravitational and cosmological sector, where naturalness reasoning is sharpest and its effective-field-theory grounding is weakest. A uniform analysis across gravitational and particle dark matter candidates shows that fine-tuning tracks the analytic structure of the abundance map, not the nature of the candidate; that the resulting classification is invariant across measure conventions while the tuning number is not; and that this decomposition instantiates informational structural realism. I situate the position against the autonomy-of-scales account, which the argument largely accepts, and against the deflationary reading, which identifies the borrowed authority but discards the structural residue.