Inference to the best explanation: Van Fraassen and the case of the ‘fifth force’