Observations of shear-wave splitting in signals from a controlled seismic source have never previously been used to parameterise a rockmass in a mine environment. In developing the necessary processing and modelling techniques and interpreting the final results, I demonstrate the usefulness of such a controlled-source shear-wave experiment to parameterise non-destructively a granite rockmass in situ and monitor non-destructively the effects of excavation. A Schmidt Hammer is used to pulse the free end of a nylon rod inserted down a 40cm borehole to create the seismic signals. I show the resulting seismograms to be highly reproducible, with cross-correlation coefficients of 0.90 or greater resulting for repeated raypaths, and that the shear-wave motion produced matches that due to an infinitesimal, isotropic, directional point source. I demonstrate that the use of multiple source orientations of a known source for repeated raypaths greatly improves the reliability of picking shear-wave polarisations and time delays. Thus the use of multiple source orientations of a known source is highly desirable in any controlled source shear-wave experiment. I demonstrate the use of cross-correlation to be effective in detecting temporal changes and that particle motions need only be displayed on the planes perpendicular to the source-receiver directions when interpreting changes in shear-wave particle motions, which is convenient for large datasets. I identify temporal changes due to the advancement of the zone of excavation disturbance, which suggests that Extensive-Dilatancy Anisotropy is at least partially responsible for the in situ anisotropy. An increase in the strength of anisotropy suggests that excavation creates an anisotropic fabric of dry cracks with orientations governed by either the in situ stress field or mineral alignment. These results suggest that shear waves may be used to remotely monitor a rockmass.