Accountability in Science

by Benjamin Kuipers

It is a tragedy when a young and promising scientist throws away a productive scientific career by committing fraud. The progress of science depends on the intellectual integrity of scientists. If an individual shows that his or her integrity cannot be trusted, their opportunity to contribute to the body of scientific knowledge is gone.

But the body of science itself is well protected, by its very nature. A scientific paper is not like a jewel, that might be counterfeit without detection. Rather, it is like a brick that goes into a wall, serving as part of a larger structure. Each brick is tested when the next brick is placed on top of it.

In science, what it means for a result to be important is that other people can build upon it. If I am building on someone else's result, I must first understand it thoroughly, and then understand how my plans relate to the earlier result. If possible, I may even replicate the earlier work, to make sure I get the same result. Whether or not I do an exact replication, if the results of my work conflict with the previous paper, it raises questions that require further investigation.

Deliberate fraud is extremely rare in science, I believe, but honest error is not. Scientific procedures are carefully designed to help scientists catch their own errors, those of others, and those resulting from random chance coming out badly. These same tests catch fraud.

What this means is that if a result is important, fraud will be caught. If the result is important, someone will build on it, so they will test it, so error and fraud will be detected. The only way to get away with fraud is with a result so unimportant that no one wants to build on it. If it later becomes important, it will be tested and caught.

And that's exactly what happened in this case. The graduate student, XXX, needed to build his own work on the results published jointly with YYY. He requested the experimental notes, so he could test the foundations he would build upon. She refused to deliver them, and the fraud was uncovered. YYY's career is ruined, and XXX has lost years of work, but the body of scientific knowledge, one of the great assets of the human race, has protected itself.

Just as bankruptcy is the way that free-market capitalism protects the economy from inefficient production and unwanted products, this is how science protects itself from bad results, whether deliberate or not.

The system works, and it works well.

Written 11/18/2001, in response to news reports of a case of scientific fraud. The names have been removed.