One of the basic unresolved issues regarding plate tectonics is the degree of coupling between surface plates and convecting mantle below. Are the plates effectively decoupled from the mantle flow field by a low viscosity asthenosphere, or are they strongly coupled to mantle flow? While these two views are essentially incompatible, they both do a reasonably good job of predicting the motions of the surface plates, and cannot therefore be distinguished on this basis. The significant distinguishing feature for these models is the magnitude of basal shear stress that is applied to the base of the plate. While it is difficult to measure this stress directly, it is possible, in principle, to measure the corresponding deformation of the plate through observations of seismic anisotropy and to infer stress. Here we focus on the Canadian Shield, for which we expect strong plate-mantle interaction. We show that seismic anisotropy can be used to constrain the magnitude of the stress level applied to the base of the plate, and to document the level of interaction between tectonic plates and the mantle below.