A scientific paper reads as a clean progression from question to result. And yet, research almost never happens that way. Experiments fail, conversations redirect projects, and none of it survives ...