Hills in nature are wrapped in a blanket of grains. Over geologic time, soil upon these hills slowly creep, but appear immobile and frozen at a glance. In this dissertation, I test this perception directly with experiments wherein grains are poured on a table and left to sit. Surprisingly, grains are not still - but slowly creep under gravity alone, with rates that are comparable to hillslope soils. I demonstrate that additional features of complexity found in nature - variable grain shapes and sizes, boundary roughness, and the presence of weak disturbances - do not extinguish creeping motions, but rather paint a picture of a slowly relaxing glass. This phenomenology is resilient and robust, and offers a new view of soil creep as the deformation of an amorphous solid, poked and prodded by disturbances in the environment that maintain soil in a perpetually fragile state. At a conceptual level, this work seeks a transfer of concept between soft-matter physics and geophysical contexts, and also provides new directions in the treatment and interpretation of field data, preparing the ground for future work examining dynamics upon Earth's surface.