A central paradox of the global monsoon record involves reported decreases in rainfall over land during an era in which the global hydrologic cycle is both expected and observed to intensify. It is within this context that this work develops a physical basis for both interpreting the observed record and anticipating changes in the monsoons in a warming climate while bolstering the concept of the global monsoon in the context of shared feedbacks. The global-land monsoon record across multiple reanalyses is first assessed. Trends that in other studies have been taken as real are shown to likely be spurious as a result of changes in the assimilated data streams both prior to and during the satellite era. Nonetheless, based on satellite estimates, robust increases in monsoon rainfall over ocean do exist and a physical basis for this land–ocean contrast remains lacking.