In his review of my book about Jan Hendrik Schön's fraud at Bell Laboratories, Plastic Fantastic: How the Biggest Fraud in Physics Shook the Scientific World, Martin Blume argues that the vigilance of whistleblowers is part of the natural corrective process of science ('Keeping up scientific standards' Nature 459, 645–646; 2009). I disagree.
Blume's analysis fails to explain why whistleblowers so often find themselves working outside ordinary channels. In science today, the activities of searching for manipulated data, pursuing charges against colleagues and investigating others for misconduct are considered extra to normal scientific activity. Many suffer for speaking up, damaging their reputations. The fact that, in this rare case, the establishment did come around to the whistleblowers' position is not a reason to play down the risks they face, or the initiative and imagination needed to work around or take on reluctant scientific institutions.
Also, the correction of fraud often fails to happen through other recognized channels. Blume mentions the attempted replication of experiments by other laboratories and peer review, which were seriously wanting in the Schön case. Schön was only exposed when information about data irregularities previously noticed inside Bell Labs was passed along a chain of concerned scientists to independent external researchers. Whistleblowers only became involved after management failed to act effectively on internal concerns.
Science today can claim ownership of some corrective processes, including criticism of unrealistic research findings. Whistleblowing, sadly, is not among them.