For many decades, ecologists have asked what prevents herbivores from consuming most of the plant biomass in terrestrial ecosystems, or "Why is the world green?" Here I ask the analogous question for detritivores: what prevents them from degrading most of the organic material in soils, or "Why is the ground brown?" For fresh plant detritus, constraints on decomposition closely parallel constraints on herbivory: both herbivore and decomposer populations may be controlled by plant tissue chemistry from the bottom up and predators from the top down. However, the majority of soil carbon is not plant litter but carbon that has been consumed by detritivores and reprocessed into humic compounds with complex and random chemical structures. This carbon persists mainly because the chemical properties of humic compounds and interactions with soil minerals constrain decomposition by extracellular enzymes in soil. Other constraints on decomposers, such as nutrient limitation of enzyme production and competition with opportunistic microbes, also contribute to brown ground. Ultimately, the oldest soil carbon persists via transformation into complex molecules that are impervious to enzymatic attack and effectively decoupled from processing by the soil food web.