A recent conversation on GoodFellows, a podcast from the Hoover Institution, concerned Niall Ferguson’s recent book Doom, and included a comparison between military and civilian planning. This brought to mind what I have always thought was a weakness of the discipline of city planning. Good military planning, as I understand it, is based on preparing for “what if,” that is, developing different scenarios. What if this happens, or that happens? City planning is different, more like advocacy, that is, what should happen. This advocacy is based on certainties: open space is good, density is good—or bad, depending. The problem is that what planners think should happen—separation of pedestrians and cars, superblocks, megastructures—often runs into trouble when it hits the fog of life. Louis Kahn posited a city of mammoth parking structures, whereas Uber reduces the need for car ownership; Paul Rudolph imagined megastructures, whereas cities grow in unexpected directions—small increments help. Years ago, Jane Jacobs pointed out that the city was much too complicated to succumb to simple nostrums; better by far to learn from what was actually happening on the ground.