Lower crustal and upper mantle rocks commonly show localized ductile shear zones, characterized by a finer grain size than the host rocks; such strain localization implies a strain weakening process. Dynamic recrystallization accompanying dislocation creep is a major process tending to reduce grain size, and several distinct mechanisms of dynamic recrystallization have been identified; for quartz these define three 'regimes' of dislocation creep, with distinctly different microstructures as well as mechanical behavior. The low temperature, higher stress process of grain boundary bulging produces strain weakening and localization in both quartz and feldspar, whereas the higher temperature process of subgrain rotation does not. Previous work on olivine has focussed on high temperature, low stress creep, although there are indications of bulging recrystallization and strain localization at lower temperatures. The goals of this project are to quantify the constitutive behavior and piezometer relation for bulging recrystallization in quartz, and for olivine to verify the existence of this process, determine the conditions over which it operates, and quantify the strain weakening it produces. The results should allow more accurate modelling of lithospheric deformation.