For centuries, many of those who have thought and written about science have claimed that science is self-correcting, that one of the hallmarks of the scientific method is that its results are not taken to be absolutely certain, but always subject to revision and even rejection if the evidence warrants. Isaac Newton, for example, wrote in his Mathematical Principles of Natural Philosophy,
“In experimental philosophy we are to look upon propositions inferred by general induction from phenomena as accurately or very nearly true…till such time as other phenomena occur, by which they may either be made more accurate, or liable to exception.” (Rules of Reasoning in Philosophy, Rule IV)
“to draw all things else to support and agree with it. And though there be a greater number and weight of instances to be found on the other side, yet these it either neglects or despises, or else by some distinction sets aside and rejects, in order that by this great and pernicious predetermination the authority of its former conclusions may remain inviolate” (Book I, Aph. 46)
Scientific method, Bacon cautioned, should take exactly the opposite attitude.
“the really valuable part of the Fourth Rule is that which implies that a constant verification, and, if necessary, rectification, of truths discovered by induction, should go on in the scientific world. Even when the law is, or appears to be, most certainly exact and universal, it should be constantly exhibited to us afresh in the form of experience and observation.” (On the Philosophy of Discovery)
In the twentieth and twenty-first centuries, this has been almost a commonplace of the scientific method. And yet, as Carl Zimmer notes in a very illuminating, if disturbing, piece in Sunday’s New York Times, science as it is currently practiced does not lend itself well to self-correction.
The main problem, besides the one that Bacon pointed to, is that, as Zimmer puts it, “Researchers make their reputations on discovery, not de-discovery.”
For example, as I pointed to in a post here last week, there has been much controversy over research published in December suggesting that a species of bacteria had seemed to use arsenic rather than phosphorus to build its DNA, in seemingly flagrant violation of the laws of biology. In May, the journal Science, which had published the original article, ran eight critiques of the study and its results. However, none of the critics of the original study have tried to replicate the results.
One of the critics, John Helmann, a microbiologist at Cornell, explained that “I’ve got my own science to do.” The most persistant critic, Rosemary Redfield, has said that “scientifically I think trying to replicate the claimed results is a waste of time.” To exert the energy, time, and money to try to replicate experiments that, they feel, will lead to nothing more than a rejection of the original results, is not something scientists want to do.
Indeed, even if scientists do rerun an experiment, and find a problem in the original result, they often find it difficult to get published. As Zimmer notes, journal editors prefer to publish groundbreaking new research, not replications.
But, because of this, the original paper about arsenic-based life has not been retracted, and its results still stand.
If scientists today wish to tout their method as special because it is “self-correcting,” it would seem that they need to do more to ensure that the work of making corrections is considered time well-spent.