An article by Judea Pearl, available here. It's quick at 8 pages, and worth reading if you enjoy statistics (though I think people who already are familiar with the math of causality¹ will get more out of it than others²). I'll talk here about the part that I think is generally interesting: