In view of the apparent ambiguity of textural evidence, some theoretical aspects of the possibility that growing porphyroblasts deform their matrix are considered. It can be shown by a simple model that a growing grain need not work against the load pressure to displace its matrix. It is suggested that, in metamorphic rocks, all grains are in contact with a random pore phase analogous to a fluid. When a porphyroblast grows at a boundary with a mineral, not all of whose molecules will be incorporated in the porphyroblast, then surplus molecules of that mineral will be removed in the fluid. If the matrix mineral is rather insoluble, and other more soluble matrix minerals are present nearby, then the growing porphyroblast may push the insoluble mineral ahead of it into space created by the solution of soluble ones. In this way, it is possible that displacement of matrix fabrics may occur and produce textures similar to some of those described by Misch (1971).