The latitudinal diversity gradient is commonly viewed as the net product of in-situ origination and extinction, with the tropics either a generator (the tropics-as-cradle hypothesis) or an accumulator of biodiversity (the tropics-as-museum hypothesis). However, a global analysis of fossil marine bivalve genera challenges the assumption that species and higher taxa originated where they currently reside and indicates a more complex dynamic involving spatial shifts through time. Despite a sampling bias favoring temperate latitudes, the fossil record indicates that bivalve genera tend to appear first in the tropics and expand outwards without losing tropical occupancy, while high latitudes are primarily a diversity sink. When we exclude bivalve families demonstrably having a poor fossil record, a 2:1 ratio of tropical vs. high-latitude first occurrences is seen in late Miocene, Pliocene, and Pleistocene time slices (ca. 10, 5, and 1 million yrs ago respectively). The tropics are both primary diversity source and primary accumulator, suggesting that the long-standing cradle vs museum debate hinges on a false dichotomy. The two major contrarian clades --  which lack strong gradients --  are exceptions that prove the rule, with the Protobranchia showing low origination and extinction at all latitudes, and the Anomalodesmata showing extratropical diversity peaks where their origination and extinction rates reach “normal” levels and a tropical diversity low where their origination is damped. A clear picture of the dynamics shaping global diversity patterns requires the spatially explicit historical data provided by the fossil record; such approaches can provide a basis for modeling biotic responses to past and future climate changes.