An important concept in our research is the "master model" - an observational and theoretical representation of the probability of ground shaking as a function of location, time and shaking threshold. The ultimate goal is to compute, within limits, the 3-D stress tensor throughout the seismogenic part of the crust and upper mantle as a function of time and location, as well as the theoretical seismograms for any possible earthquake. The objective is to forecast earthquakes and their seismic wave and ground displacement effects. A complete specification of the stress would require perfect knowledge of the boundary conditions (including strength on faults), initial conditions, and rheological properties of the relevant volume. To compute theoretical seismograms we must know the seismic wave velocities everywhere the waves go, and for strong motion, the nonlinear properties of the near surface. While the complete master model is presently out of reach, we can develop simpler models that are widely applicable to earthquake hazard reduction. The first such model is represented by the figure below, from SCEC's "Phase II" report. The master model concept itself provides a useful focus for multidisciplinary research in geology, geodesy, seismology, and theoretical mechanics.