A quantitative-genetics formalization of latent human aptitude architecture

The concept paper "Genetic pools as substrates for investigating latent human aptitude architecture" proposed that structured synthetic populations could serve as substrates for studying the genetic architecture of latent traits: traits that are not directly measured but inferred from correlated indicators. That proposal was deliberately conceptual; it lacked the quantitative-genetics machinery that would make it testable. This paper supplies that machinery: the phenotype decomposition P = G + E and its variance partition, narrow- and broad-sense heritability, polygenic-score construction and its portability limits, the breeder's equation for predicted selection response, and the multivariate extensions needed when "aptitude" is really a bundle of correlated traits. Numerical illustrations are worked through, but they are arithmetic demonstrations of the equations, not empirical findings; every number is labeled as illustrative. The paper closes with explicit falsifiability conditions: observations that would refute, rather than merely complicate, the framework. This is a formalization exercise, not evidence that any latent aptitude trait has the genetic architecture described.