What we learn in our high school science classes about the nature of scientific progress is that it comes brick-by-brick. We learn that scientists form hypotheses, devise experiments to test them, collect data, form conclusions, around and around. Their results and conclusions are bricks that get added to the wall of science, a wall that gets built higher and higher over time with each generation of scientists.

In fact, however, that is not really how science really advances. So argued Thomas Kuhn The Structure of Scientific Revolutions (1962). Using numerous historical examples, he showed that the model I just layed out describes how science works sometimes, for some periods. Yet the real engine of science, Kuhn said, works differently: on occasion results come along that do not fit into the brick wall that everyone else is building.