The social sciences, I’ve come to believe, don’t have a coherent concept of causation. To talk about ‘causation’ we need to have a boundary on cause and effect. I reflect on what this means for studying causation.