The theory of continental drift was introduced in 1912. The mechanisms behind plate tectonics were discovered in the 1960s. Many historians of science will tell you that this theory was one of the most profound developments of the 20th century, but what do the ~300,000 papers published in Science, Nature, and PNAS tell us? I extracted 1000 themes or "topics" from this corpus with a statistical model. With no information but the papers themselves, the model chose to group together the words shown in this plot as a topic, #443. The position and brightness of each word shows the number of times that word appears in papers from each decade. But the model doesn't just count words. If just one of these words -- say, "zone" -- appears in a paper without the other words, like "fault" or "crust", it will probably be assigned to a different topic. At the same time, each paper can contain many topics. The numbers corroborate the historians: rumblings early in the 20th century presage the earthquake of the 1960s.