Einstein’s fudging . . . and other researchers behaving badly

It isn't just journalists, politicians and police officers who behave badly. On 1 August, the cognitive scientist Marc Hauser will step down from his post at Harvard to pursue "new and interesting challenges". His resignation comes after a year on leave, which followed an internal investigation into his research on the minds of primates that found him guilty of eight counts of scientific misconduct.

At Columbia University, New York, another investigation has found that a chemistry graduate student called Bengü Sezen made up a lot of her data; six published papers have been retracted. What is still not clear is whether Dalibor Sames, her professor, should have picked up on the fraud - instead, he sacked two other graduate students for lacking the skill to replicate Sezen's results.

Sezen was caught only after lab members mounted a sting operation to reveal her sabotage of colleagues' experiments. This has echoes of another case: that of the former University of Michigan researcher Vipul Bhrigu. Caught on camera last year, he confessed to interfering with the experiments of his colleague Heather Ames.

Bhrigu had been switching labels on Petri dishes and compromising samples by lacing them with antibodies. Apparently, her swift progress was making him look bad.

Scientific misconduct is more prevalent than anyone would like to think. Some of the biggest names in science - Isaac Newton, Albert Einstein, Galileo Galilei - have been guilty of questionable behaviour. Einstein cherry-picked data and fudged proofs of E=mc2 (he never managed to prove it properly); in the Principia, Newton massaged his equations to fit with the latest data. And Galileo tried to convince the pope that the earth moved around the sun by "proving" that this was what caused the tides, when everyone knew even then that it was the moon.

In 2005, the journal Nature published a report showing that a third of scientists admitted to having misbehaved in the previous three years. The cited crimes included falsifying data, ethical breaches, dropping inconvenient data points and failing to give proper credit to the work of others.

Fight club

Given all this, should we start to show a little more mistrust towards the pronouncements of science? Not at all. The interesting thing about scientific fraud is that it rarely makes waves. Its exposure never changes scientific consensus. In some ways, it provides proof of the reliability
of science. The misbehaviour is motivated either by the sheer difficulty of proving what your gut tells you to be true or by a human desire for more credit than you deserve. Nothing you do, however, can change the laws of nature. If your goal is to misrepresent the way the universe works, you can be sure that your colleagues will find you out. It's their job to poke and prod at your work and expose any flaws. They carry this out with glee: that's why it's not unknown for fist fights to break out at research seminars.

As a result, any important idea or experimental result that survives a few years in the gladiatorial arena of science is almost certainly trustworthy. Which is why, for example, we should take note of the consensus among climate science researchers about man-made global warming. It stands at 97 per cent.