Recently in The nature of science Category

I've been writing a lot about about newspapers lately. It's not because I'm particularly knowledgeable about them. It's because I'm concerned. This post is a bit different.

I pointed out a few days ago that the George Will episode provides evidence that we need news organizations and editorial pages that check their facts. Today Nick Kristof has an interesting column on experts. Here's a paragraph I found particularly striking:

The marketplace of ideas for now doesn't clear out bad pundits and bad ideas partly because there's no accountability. We trumpet our successes and ignore failures -- or else attempt to explain that the failure doesn't count because the situation changed or that we were basically right but the timing was off.

Kristof's point is one that philosophers of science, starting with Karl Popper, have made for a long time. It's necessary to look for data that would contradict your hypothesis, not data that support it. Being accountable is good. Looking for evidence that shows you're wrong is even better. That's why peer review is so important.

It's really hard to evidence that you're wrong. It's much easier to look for evidence that someone else is wrong or pushing the conclusions beyond what the data justify.

The scientific method and peer review don't guarantee that published results are right. But because peer reviewers are looking for flaws and because they are often selected because they're known to have a different view of the problem than the author, it does mean that peer reviewed science is free of obvious flaws, or at least free of flaws that weren't obvious to the authors of the paper and to three or four experts in the field.

Twenty years ago today, Tim Berners-Lee wrote a report for CERN, the Euorpean Organization for Nuclear Research, that his boss, Mike Sendall, described as "vague, but exciting." The report concerned "the management of general information about accelerators and experiments at CERN," and it "discusse[d] the problems of loss of information about complex evolving systems and derives a solution based on a distributed hypertext system."

Why do I mention nuclear research and management of information about accelerators? Well, if the name Tim Berners-Lee rings a bell, you already know that the answer has something to do with the web. What you may not know is that the web turns 20 today. That report described what would become the web, and CERN is hosted a celebration of its 20th anniversay earlier today.

I first encountered the web in 1993 or 1994 when I heard of this thing called Mosaic. There wasn't a lot available then, but I could see that it was likely to be useful, but I never imagined how ubiquitous it would become and how it would change the way that I do my work.

So if anyone ever asks you about serendipitous inventions that spring from basic research and have large practical effects, ask them how long it has been since they looked at their bank statement on the web or since they bought a book or since they downloaded a track from iTunes or since they watched a video on YouTube or since they read a newspaper or magazine in their favorite web browser or....

Larry Moran posted this video about a month ago, but it's very good and worth bringing a little attention to again.

We've all seen a movie (or Numb3rs) where suddenly a scientist says "Eureka!" and a great discovery is made. Those of us who are scientists know that those "Eureka!" moments are rare,¹, and so does The Economist. In writing about the Large Hadron Collider and the discoveries that could be made there, they had this to say:

Scientific discoveries are only occasionally eureka moments. More often, the data have to be collected, reviewed, analysed statistically, found wanting, collected again and re-analysed. Eventually, if all has gone well, a clear result will emerge. It then has to be written up, reviewed by critical peers and, if it passes review, published in a scientific journal.

Scientists understand the world first by noticing regularities and second by explaining them. Linnaeus noticed that living things could be grouped into hierarchical categories more than a century before Darwin provided the explanation. Mendeleev noticed that elements could be grouped into a periodic table more than a century before the atomic theory explained why.

One of the challenges we always face is figuring out what patterns make sense. Take a bunch of nucleotide sequence data from a bunch of different nuclear genes within a large population of an outcrossing species, throw it into your favorite phylogeny program, and you'll get out a tree, one tree¹ -- even if each gene has a different evolutionary history.² Or take a bunch of data from a single gene and a bunch of different populations and throw it into the same programs, and you'll get out a tree -- even if the populations show a linear cline or isolation by distance.

Wouldn't it be great if you could throw your data into a program and have it figure out whether a tree is the best way to structure your data or if some linear order or a dominance hierarchy or something else made more sense? Well, hang on to your hats. There's a recent paper suggesting that it might just be possible.

(Thanks to Orac for pointing this out. Image from bighugelabs.com)

No, this isn't about ScienceDebate2008. Nor is it about recent challenges to the integrity of science. It's about drawing a lesson from recent events that illustrates an important wy in which the process of reasoning science is different from our everyday reasoning, a difference that is often poorly understood.

If you're a Democrat, your candidate won in Wednesday night's presidential debate -- that was obvious, and most neutral observers would recognize that. But the other candidate issued appalling distortions, and the news commentary afterward was shamefully biased.

...

To understand your feelings about Wednesday night's debate, consider the Dartmouth-Princeton football game in 1951. That bitterly fought contest was the subject of a landmark study about how our biases shape our understanding of reality.

Psychologists showed a film clip of the football game to groups of students at each college and asked them to act as unbiased referees and note every instance of cheating. The results were striking. Each group, watching the same clip, was convinced that the other side had cheated worse – and this was not deliberate bias or just for show.

“Their eyes were taking in the same game, but their brains seemed to be processing the events in two distinct ways,” Farhad Manjoo writes in his terrific new book, “True Enough: Learning to Live in a Post-Fact Society.” (Nicholas Kristof, Divided they fall, The New York Times, 17 April 2008)

From the mid-1970s to the mid-1990s Congress had something called the Office of Technology Assessment.

The congressional Office of Technology Assessment (OTA) closed its doors September 29, 1995. For 23 years, the nonpartisan analytical agency assisted Congress with the complex and highly technical issues that increasingly affect our society. (source)

OTA issued reports on topics ranging from addiction, aging, and agricultural technology to waste management, and women's health (see the full list at http://www.princeton.edu/~ota/ns20/topic_f.html). Those of you who've been around for awhile will recognize 2005 as the first year of the “Gingrich revolution.” The Republicans in charge of Congress apparently decided that they didn't need non-partisan advice on complicated technical issues.

Wrong! As I've written elsewhere, “Science can describe the outcomes associated with different policy choices, but the choice between those outcomes is determined by what we value” (source). But to make the right policy choices, policy makers need to know the consequences of their policy choices.

Mark Hoofnagle started a campaign a little over a week ago to bring back the OTA. When I checked this morning I found links to 24 blogs that have joined the campaign. I may be slow on the uptake, but I'm in now. Expect posts periodically describing some of OTA's past accomplishments to provide ammunition for the campaign.

Timothy Sandefur and Ed Brayton have long posts on an article published in the law review of the Chapman University law school:

Trask, S. W. 2006. Evolution, science, and ideology: why the establishment clause requires neutrality in science classes. Chapman Law Review 10:359.

The article isn't available on line, and I haven't seen a hard copy, so you'll have to refer to the posts above for a detailed analysis. I just want to point out that Sandefur gets the title of his post right, “All epistemologies are not created equal.” It's right because, as Sandefur notes, the scientific method has proven itself unequaled as a method of learning about the observable world. And it's right because Sandefur correctly locates the argument about the status of revelation, religious belief, and other ways of knowing as an argument about epistemology, i.e., an argument in philosophy, not science.

In that sense, Sandefur's argument goes even further than is necessary. He argues that science and reason are the only legitimate source of knowledge about the world. But to show that creationism doesn't belong in science classes we don't need to deny the legitimacy of religious belief, and we don't need to claim that science is the only legitimate source of knowledge about the world. We only need to show that creationism doesn't follow the norms and practices of science, a task that's been repeated innumerable times.