A few years ago, a philosopher and a sociologist sought my advice on a paper they were working on. After they had explained the nature of their troubles, I asked them if they would let me assign a short, hour-and-a-half writing task them for them. They accepted. The task required them to each write — without discussing it in advance — a three-paragraph introduction for the paper they were struggling with. They were to go to their corners and spend a morning composing a paragraph about the world their research addressed, another about the science they were applying to the problem, and, finally, a paragraph about the paper itself. They were to write the first three of my “five easy paragraphs”, devoting no more than half an hour to each. Then, without showing each other what they had written, they were to send their work to me. After a few days, I received their introductions.

You can probably guess what happened next. It was now easy for me to tell them why this paper was difficult to write: they were writing two different papers, framed