Friday, January 24, 2014

Different Fields of Science Do Not Have Different Standards of Proof

Back in 2010, a team of researchers claimed to have discovered a life form that could exist without phosphorus, instead using arsenic in such key molecules as DNA, RNA and proteins. The research was published in Science and almost immediately was under intense scrutiny from chemists and biologists. Somewhat surprisingly, physicists and geologists didn't understand what the fuss was about. The work has since then largely been discredited.

Last October, a trio of scientists wrote an article in which they argued that different fields of science have different "proofs". The article is behind a pay-per-view firewall (and I can't access it), but one of the authors has written an open access summary in Chemistry World. In both articles, the scientists argue that this wide dichotomy in acceptance and rejection of the results came about each field having its own standards of "proof". From the Chemistry World article:
"Proof for a chemist requires different data and arguments than proof for a physicist."

Bull cookies.

Science has long known that it can't "prove" anything. It can only disprove something. Einstein was asked about proving his Theory of Relativity and said, "No amount of experimentation can ever prove me right; a single experiment can prove me wrong." And yet here we have a scientist arguing that you can "prove" something.

What is even worse about the idea of different fields having different standards of proof is what it then implies about peer-review. Such a comment necessarily means that peer-review (pre-publication, post-publication or both) somehow puts a gold star on research, a stamp of approval that it is "proof".

Hog wash.

Much of the public believes that peer-reviewing means that the research and conclusions are true, an idea that many (including me) have been fighting hard to correct. And now we have one of our own reinforcing that very bad idea.

Science is built up from contributions made by teams and individuals. In some cases, the contributions are found out to be wrong, whether it is because they contradict the rest of science or because they cannot be replicated. The wrong contributions are removed and science goes on. In some cases, the corrections come quickly, in other cases it may take decades or even centuries, but science always builds an internally consistent, self-correcting house. And it is one house. There is not a special house for biology and another one for chemistry and a third for physics. One house, one science.

Science will correct itself as it did in this case. But this was because conclusions based on the evidence were contradicted, not because chemists and biologists had different standards of proof. Physicists and geologist can understand what went wrong in the experiments without having to accept new standards of proof. The whole of idea of different standards of proof exists in the courthouse, but not in science.

Previous Years
January 24, 2012 - Confusion over "Vinyl"

January 24, 2011 - How about a Deborah Number for Durability?

No comments: