Dediscovery: My new essay for a new section of the New York Times

By Carl Zimmer | June 27, 2011 10:41 am

In the late 1800s, prominent astronomers declared that Mars was criss-crossed by canals–evidence, they declared, of an advanced civilization. But in the early 1900s, astronomers gazed through more powerful telescopes and discovered that the canals were mirages.

The astronomer Percival Lowell, who had become the leading champion of the canals, scoffed at the new findings He declared that the criticism came “solely from those who without experience find it hard to believe or from lack of suitable conditions find it impossible to see.”

Although the new evidence led many astronomers to abandon Lowell’s position, he never retracted his claim. It wasn’t until five decades after his death in 1916 that space probes finally went into orbit around Mars and sent back close-up pictures of a canal-free Red Planet.

I’ve always been fascinated by the way science casts aside bad ideas. For most of us, it’s easy to assume that science shakes them off quickly, but the truth is that it can take quite a while for the process to play out. Recently I was invited to contribute a piece to the new “Sunday Review” section of the New York Times, which just debuted this week. I wrote an essay on this phenomenon, which has been dubbed  “de-discovery.” I drew on three recent examples of high-profile research that many other scientists have declared to be wrong–arsenic life, clairvoyance, and a link from chronic fatigue syndrome to a virus called XMRV.

To keep my essay from exploding into a novella, I had to limit myself to just these three examples–but I could have picked many others. You just need to check out a blog like Retraction Watch to see how important this part of the scientific process is today. The first draft of my essay actually started out with a fourth example, which I decided to cut it in the end. It’s a peculiar case of a de-discovery of a de-discovery.

In 1981, the late Harvard paleontologist Stephen Jay Gould published an influential book about racism and science, called The Mismeasure of Man. Gould argued that social influences could lead scientists to misinterpret their data to suit their beliefs about European superiority. One of his key examples was the work of a nineteenth century anthropologist named Samuel George Morton.

Morton collected 1,000 human skulls from around the world and measured the size of their brain cavities with seeds or lead shot. Gould re-analyzed Morton’s data and published his results in 1978 in the journal Science. He declared that Morton fudged his measurements to ensure that Caucasians would end up with the biggest brains.

In 2000, a freshman at the University of Pennsylvania named Jason Lewis started to measure Morton’s skulls for a research project of his own. He was interested in the ways different human populations adapt to different climates—including changes in the shapes of their skulls. It was then that Lewis learned from his advisors about the controversies swirling around the skulls. (He was born a year after The Mismeasure of Man was published.)

As Lewis carried out his own measurements, he gradually realized that Gould had been wrong. He then set out to systematically investigate the matter—taking three years to measure Morton’s skulls, and then another five years to work through Gould’s claims.

Lewis, who just finished earning his Ph.D at Stanford University, wrote up the results with his colleagues and submitted a paper in 2008 to the journal Current Anthropology, which had published a less detailed critique of Gould’s paper in the 1980s. The journal rejected Lewis’s paper, eventually informing him that it was not important enough.

The researchers had better luck with PLOS Biology, which published their paper earlier this month. Lewis and his colleagues presented evidence that Morton did not bias his findings at all. Instead, the researchers conclude, it was Gould who used shoddy statistics. There are many sound scientific reasons to reject racist views of human biology, they argue, but an unfair trashing of Morton’s research isn’t one of them.

“Our analysis of Gould’s claims reveals that most of Gould’s criticisms are poorly supported or falsified,” they write.

When I was researching my essay, I asked Lewis about what he thought of science’s self-correcting process now that he’s finally done with his exploration of Gould and Morton. He has decidedly mixed feelings.

“We can come back thirty years later and get the story straight,” he told me. “But it takes thirty years.”

As I write in my essay in the Times, there are certainly ways to make dediscovery a smoother, faster process. But in an age of instant viral communication, I think we’re going to remain frustrated by inescapable lags.

[Image: Wikipedia. Thanks to folks on Twitter for pointing me to Martian canals as a textbook case of slow dediscovery]


Comments (6)

  1. Christiana Leonard

    My husband and I have followed your work ever since I discovered your wonderful account of Thomas Willis and the Oxford group. I was soooooo happy to read your article yesterday. It was absolutely shocking to learn that a major journal would have a policy of not publishing replications. I went into science after taking philosophy of science courses that taught me that the great advantage of the scientific method was that consensus could be used to discover truth. Individual observations can be faulty but if there are lots of observations (no matter how seemingly subjective) that agree, one can gradually become more and more confident. I was awestruck by this insight and have never regretted my choice. Since I left animal research and entered the human imaging field in 1990, I have devoted myself to the lonely task of trying to replicate findings — mine and others. In schizophrenia research, one article stated that findings disappear when you double the sample. That has been my experience, too, although I have found a few relationships that could be replicated in new samples. It is really shocking how few scientists in human research think that replication is essential or even important. Considering how hard it is to do, maybe that is no surprise. In my field, many findings that are patently untrue are quoted and cited over and over again as divine truth, “Dylexics have migration abnormalities, ” “The size of the brain is unrelated to intelligence,” for example. Your column made me wonder if you were planning to write a book on this?

  2. 220mya

    I was going to leave a snarky comment about how space probes did not circle Mars in 1916. Then I realize it depends on how you read that sentence. In any case, might want to re-order things so people don’t get confused as to whether 1916 refers to Lowell’s death or the space probes.

  3. Jack

    I did enjoy reading your piece both here and in the NYT.

    One point though, surely ‘de-discovery’ only happens in hindsight. I mean as a scientist one wouldn’t enter into research to disprove a theory would one? That would be subjective and biased.

    Rather one would set out to prove a theory, to replicate or not and then publish one’s findings, which might then be taken as ‘de-discovery’ by others; until someone else comes along and ‘re-discovers’ perhaps!

    Who decides when a theory is debunked anyway? Is it really only decided by the publication process and peer-reviewers? When does ‘science’ actually say – ‘Right that’s proved! NEXT!’

    It hardly sounds scientific to me – this whole process. Maybe there should be a separate system for attempts at replication? Maybe there is?

    The current system hardly warms one to the process by which, for example, vaccine production might be stopped or mass blood-screening protocols halted, all because one accepted theory – presumably replicated repeatedly – is suddenly ‘de-discovered’.

    I mean just how much power – including public hysteria – do these papers that claim ‘eureka!’ actually have in today’s internet world? It seems that as soon as one is published that claims – so-and-so – it can and does create a tsunami effect.

    And then all successive attempts at ‘replication’ which fail to – or support even – the original findings – if not the original methods precisely [rolls eyes] – attract no end of criticism from patients, professionals and media.

    Not a very comforting position at all this system of publication. Seems all so ‘hit and miss’ to me I am afraid.Then what do I know? A former Private Banking Investment Manager :)

  4. Mike

    I read an account once (can’t remember the source) that although the “scientific model” has scientists persuaded by evidence, what really happens is that young scientists adopt the new approach and older scientists cling to what they’ve always believed true. But since they’re not training future researchers (grad students are smart enough to get with the new program) the beliefs die out with them. In a generation the old belief is seem as nothing but a quaint story. This seems to dovetail nicely with the 30 years timeframe mentioned in your article.

  5. meir

    I think that the problem is that science is not done by scientists – it’s done by PEOPLE. And people, as some of us may know, are highly unreliable/ It should probably be more of a surprise to us that science makes any progress at all, don’t you think?


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

The Loom

A blog about life, past and future. Written by DISCOVER contributing editor and columnist Carl Zimmer.

About Carl Zimmer

Carl Zimmer writes about science regularly for The New York Times and magazines such as DISCOVER, which also hosts his blog, The LoomHe is the author of 12 books, the most recent of which is Science Ink: Tattoos of the Science Obsessed.


See More

Collapse bottom bar