Five decades of attention research have produced five major but isolated theories. I propose the Magnetic Field Model of Attention (MFA), a unified framework in which attentional intensity decays as F = S/rα. The exponent α = 2 is derived independently from geometric flux conservation, neural wiring-cost optimization, and Amari-type neural field equations. Six field equations formalize attentional gradients, capture, energy conservation, superposition, task-switch inertia, and priming. Five parameters span the ADHD-to-hyperfocus continuum. Reanalysis of published gradient data using AIC model selection favors power-law over Gaussian decay (ΔAIC = 6.6); reanalysis of load data confirms continuous over binary resource allocation (ΔAIC > 10 against step function). Eight falsifiable predictions are proposed.
Cora Zeng (Sat,) studied this question.