One of the most important questions in planetary-tectonics research is whether endogenic stresses, such as mantle convection on Earth, play a significant role in governing global tectonic evolution of a solar-system body. In this study, we investigate the relative importance of endogenic vs. exogenic stresses in controlling the spatial location and temporal variation of active ice-shell deformation on Enceladus, which is expressed by cyclic plume eruptions along active fault zones (i.e., the tiger stripes). Although the variation of the eruption flux on Enceladus follows the periodicity of the diurnal tide, it remains unclear why there is a consistent phase delay of the observed peak eruption when compared to the predicted peak tidal stress. Here, we explore whether endogenic stresses in the ice shell are capable of explaining this observed phase delay. To achieve this goal, we performed geologic mapping along the tiger-stripe faults that host the erupting plumes. Using the fault kinematics established from our mapping, we determine the general stress state (i.e., the principal-stress directions) along the tiger-stripe faults. This knowledge in turn forms the basis for inferring the most likely plume-eruption mechanism. Our mapping shows that the tiger-stripe fractures are not tensile cracks but are instead left-slip fault zones locally displaying extensional fissures. This insight leads to a hypothesis that strike-slip faults and their local tensile cracks experience simultaneous shear and tensile failure, and that the tensional opening reaches maximum at the time of the peak plume flux. We quantified this hypothesis using a stress decomposition model that assesses (1) the relative importance in magnitude between the tectonic stress and tidal stress exerted on the tiger-stripe faults and (2) the role of ice-shell properties such the shear and tensile strengths and ice-shell thickness in controlling the eruption phase delay. Using laboratory-determined ice strengths and the best estimate of the ice-shell thickness at the South Polar Terrain of Enceladus, which hosts the tiger-stripe faults, our model results indicate that the endogenic tectonic stress is comparable in magnitude to the tidal stress. Although we cannot rule out warm-ice convection, true polar wander, and non-synchronous rotation as causes of endogenic stresses, the large variation in ice-shell thickness makes the lateral gravitational-potential gradient the most plausible source of the endogenic stress required by our model results.