LAUSR

1 _ _   cmplx Math cmplx Geol . (1) This statement summarizes two end-member cases of modeling that can be illustrated with examples from mechanics of faulting. On one hand, highly complex multi physics/chemistry processes in heterogeneous crustal-scale structures are often analyzed with back-of-theenvelope approximations and simple terms such as dislocations on a planar surface in a homogeneous solid. On the other hand, far simpler engineered rock samples in laboratory experiments under controlled conditions are often analyzed with elaborate mathematical models of complex constitutive laws having many parameters and variables. As noted by the statistician George Box, "all models are wrong but some are useful". The question, keeping in mind Sibson's observation, is which models are useful in what circumstances. The answer depends on whether one is interested in engineering applications or scientific understanding.