Suggestions that the 2008 research assessment exercise may be altered or replaced are ludicrous. To do so would offer endless opportunities for legal challenge about outcomes, because of procedural changes made after publication of the criteria. But the most damaging structural suggestion in the debate is that research assessment may be differentiated among disciplines. Considerable effort has been invested in creating an RAE with common procedures but appropriate, accepted differentiation among disciplines. This near-level playing field offers disciplines even chances to perform well overall. How can this occur in a procedural free-for-all?
Interdisciplinary research will also suffer. The RAE criteria at least show intent to construct mechanisms to assess research combining approaches from different traditions. That the present debate was triggered by the Treasury is ironic. In 2004, the Treasury stated: "We need to enhance a culture of multidisciplinary research... and provide infrastructure and funding... to support it. Over the next decade, many of the grand challenges in research will occupy interfaces between...
disciplines." Differentiating the assessment practices seems destined to drive a wedge between disciplines that should be encouraged to work together on these challenges.