On being wrong – or not completely right

Stephen J Gould is one of my heroes, one of my favourite scientists. One of his best books, in my opinion, is The Mismeasure of Man – which was written in part to challenge the use of IQ tests, phrenology and other ’empirical methods’ to type people. Typing people in this manner is a dangerous game and Gould in his book very deftly explains how statistics can be misinterpreted and importantly cautions against establishing false causal relationships (such as Baroness Greenfield’s dictum ‘I point to the increase in autism and I point to internet use’).

Gould in his day job was an evolutionary biologist who disagreed with much that was written in Dawkins’ tome ‘The Selfish Gene’; the mechanism of evolution, the presence of gaps in the fossil record, the gene being the fundamental unit of evolution (Gould thought it was the species).

In recent years, after his death, there have been several articles and blogs which attack Gould for being a fraudster who claim he ‘committed the moral equivalent of deliberate scientific fraud.’ Gould allegedly stole someone else’s idea and if this weren’t bad enough it is asserted that he had no credibility amongst other evolutionary biologists (such as Richard Dawkins and EO Wilson). This seems a bit far-fetched to me. Just because he disagreed with Dawkins (and Wilson) doesn’t mean he is not credible and why accuse Gould of ‘stealing ideas’ if not Dawkins whos ideas are reflected by EO Wilson (who predates Dawkins).

Gould also claimed that bias is inevitable in science and that we must be aware of this. Gould wasn’t a conspiracy theory kind of guy, I would argue, he just had a more gentle point about unconscious bias. In July last year, a group of scientists remeasured some skull volumes Gould had originally measured to point out bias in an older phrenology investigation. They came to the conclusion Gould was wrong and the original researcher (who was openly biased) was correct.

The title of this article (and tag line) in ‘New Scientist’ is:
Gould’s skulls: Is bias inevitable in science? Stephen Jay Gould claimed unconscious bias could affect even seemingly objective scientific measurements. Not so

Despite its enticing headline, if you read the article itself, it doesn’t appear that the authors are vehemently disagreeing with Gould but are rather saying that Gould might in fact be biased. I can’t help but thinking this might make Gould happy. He was the first to admit he himself was biased.

Does any of this make Gould a ‘bad’ or ‘fraudulent’ scientist? No, not at all. Let me tell you why.

Dawkins and Gould disagreed about some of the fundamental drivers of evolution. They both came up with a hypothesis of how evolution occurs. Both of these hypotheses are consistent with a set of available data. A scientific theory is an educated guess which is consistent with a set of observable data. Both Gould’s hypotheses AND Dawkins’ hypotheses are perfectly valid scientific guesses – and neither has YET been proven right or wrong. A particular ‘selfish gene’ which can prove or disprove Dawkins, for instance, has not been discovered. This is how the scientific method works.

Secondly, did Gould steal his ideas? This is impossible to prove and science is often, for lack of a better word, repetitive and has much to do with the current Zeitgeist. If you read the literature about a subject any subject, often people are saying similar things in many different ways. ‘Who said it first’ is difficult to establish, especially when ‘who said it first’ is not well known.

Both opposing hypotheses and replication of ideas happen every day, in fact I would say this is part of the routine of practising science

My old post-doc supervisor sent me a paper he is about to submit last week, asking me what I thought about it. In this paper, my old boss has partially disproven a hypothesis that arose from some work he and I did about 7 years ago. He came to his current conclusions by measuring some NEW data which weren’t available at the time of our first publication. The upshot is a part of what I hypothesized was happening in my paper is ‘wrong’ while a part of my hypothesis remains correct. I thought this new paper was a great paper, it furthers our understanding and modifies our original hypothesis. This is how science works at its best.

Also this week my post-doc found a new paper which fully supports a hypothesis I made a few years ago. The authors didn’t cite my original work. Does this mean they stole my idea? No. It mostly likely means that they haven’t seen my publication or read my original hypothesis. This is not abnormal, there is a giant amount of literature out there and scientists don’t all read the same journals.

Often things are much milder than they seem and science is about getting it wrong sometimes.