Silicon carbide was subjected to dynamic deformation in (a) a compression Hopkinson-Kolsky bar (compressive stresses of 5 GPa), and (b) high-velocity impact under confinement (compressive stresses of 19 GPa) by a cylindrical tungsten alloy projectile, and (c) in the radial collapse of pre-fragmented or granular ceramic in a thick-walled cylinder geometry. Shear localization was observed to accommodate the large permanent deformation in the radial collapse geometry. Localization occurred by two different micromechanical modes: for larger fragments, the material within the shear bands was finely pulverized; for smaller particles, the material within the shear zones exhibited the effects of heating and subsequent sintering For specimens from the Hopkinson bar and confined impact experiments, considerable evidence of plastic deformation, as dislocations and stacking faults, was observed in the recovered fragments. Profuse dislocation activity was observed in the frontal layer in the specimen recovered from the projectile impact. The formation of this layer is enabled by the high lateral confinement. Plastic deformation is consistent with an analysis based on the competition between crack growth under compression and dislocation generation. Different mechanisms are discussed for the initiation of fracture; (a) dilatant cracks induced by mismatch in the effective elastic moduli between two adjacent grains, leading to internal tensile stresses and creating transgranular fracture; (b) Zener-Stroh cracks nucleated by the piled up dislocations along grain boundaries, and resulting in intergranular fracture; (c) cracks due to existing flaws; (d) stress concentrations due to twinning and stacking faults. Heat transfer calculations are carried out that show: (a) the intact layer between the Mescall zone and the impact interface is not due to heating and sintering, but due to unique state of confinement provided by the uniaxial strain state, and (b) that the temperatures wit lain the shear zones can reach values of 2,500 K, consistent with the observation of sintering.