Progress and fashion

A popular view of scientific research is that it is rather like a detective story. A crime is committed. The unimaginative police inspector puts a conventional interpretation on the more obvious clues and makes an arrest. Then along comes some eccentric who views the crime as an intellectual puzzle, worries about apparent trivia that fail to hang together, and, as one suspect after another is eliminated, comes up with the true answer, so that the innocent prisoner is freed and the murderer is handed over for punishment.

Up to a point, the analogy holds. At the level of trying to work out why a specific effect occurs, research can indeed seem like detection. Certainly, it involves an element of fussing over facts that seem not to fit with each other, and a preparedness to challenge conventional solutions if they do not seem quite conclusive. A large part of research design involves playing around with alternative possibilities, until one can crystallize one's reasoning along such lines as: ‘if theory A is correct, then, under these conditions, this should happen. If theory B is correct, then, something else should happen.’ One of the things that can be most fun about doing research is seeing some of these ‘if-then’ connections one imagined turn into actuality in the response of one's subjects.

The analogy starts to fail, however, as soon as one starts to move away from the confines of a specific paradigm to a choice of broader theoretical perspectives.