It is proposed to study small-scale velocity differences in open and closed turbulent flows. Of special interest is the cross-over from the chaotic state (few degrees of freedom) to the fully turbulent state, where "cascade" ideas of turbulence may be applied. The turbulence will be probed by conventional techniques, such as laser Doppler velocimetry, and also by homodyne light scattering, which permits the measurement of velocity differences without making the "frozen-turbulence" assumption. Also to be studied are: the dynamics of phase-separation of fluids and liquid crystals contained in porous media; and nucleation in fluid mixtures under uniform shear. The common thread running through all of this work is that of systems driven far from equilibrium. %%% The research deals with "self-similar" phenomena which have the property that they look the same when viewed on all scales of magnification. Turbulent fluids exhibit this self-similarity, which will be studied by light scattering. Another self-similar occurrence is the motion of fluids constrained within porous media (e.g., oil in sand). The goal is to understand the origin of this self-similarity.