Every once in a while I come across a behavioral task (typically in rodents or monkeys) that is far more interesting than the accompanying analysis of neural correlates. It would be nice to read a historical survey of task design: it seems underappreciated.