My thinking is, when a fairly non-periodic, non-polar compound is subjected to an electric field, dipoles emerge, and although they're all in the same direction, since the arangement of atoms in the substance isn't that regular, then the electric field inside will be non-uniform (on the microscopic scale). If there are small changes in the electirc field due to neighboring dipoles, then the dipoles will feel forces, thus causing more net movement and temperature. Seemingly, in this case the applied electric field might raise the temperature. But the amount of temerature increase must be dependant on the polarizability. The polarizability of substances decreases as temperature increases, so there must be an equilibrium point. So this tells me that the polarizability is like an extra "storage tank" of temerature of the substance, hinting towards a relation with heat capacity.