We introduce a new model of geographic mobility that maps the underlyi
ng contours of sociocultural space after purging the confounding effec
ts of distance, inertia, contiguity, and population size. In conventio
nal models of geographic mobility, these confounding effects are inten
tionally conflated with sociocultural forces, since the objective is t
o fashion an abstract social physics of mobility that is based on a un
itary and all-encompassing distance function. We argue that known feat
ures of the social landscape should be entered explicitly into mobilit
y specifications. This approach yields new insights into (1) the net h
olding power of regions, subregions, and states; (2) the effects of ma
cro-level variables on propensities for immobility and exchange; and (
3) the influence of seemingly minor variations in the shape of regions
on gross patterns of immobility and exchange.