Tuesday, July 22, 2008

Tiny Bubbles

Bubble Fusion Researcher Charged with Misconduct: Scientific American

This is s pretty good article on the Taleyarkhan scandal. Fortunately SciAM gets it right and identifies Taleyarkhan as an engineer, rather than as a physicist. This is what ASEE did in it's daily news digest "First Bell" yesterday, running a report from the Los Angeles Times that made the same mistake. I sent an email to the LA Times reporter who wrote the story, but no reply - hopefully mine was not the only email he received.

I can understand an engineering organization like ASEE not wanting to identify this guy as an engineer. And I do not want to pretend that improper professional behavior has not happened in physics. We recall the famous cases of recent years, the fraudulent claim of discovery of element 118 (which was later legitimately discovered) and the Schon case at Bell Labs. These two cases caused the APS to go into full frantic mode and pushing professional education in physics curriculum and a number of other remedies.

But there are some differences here. In the Ninov case (element 118) there was every reason to believe that the result was plausible - indeed the element was eventually seen, including by some of the researchers who had worked with Ninov. The Schon results, although remarkable, seemed plausible as well. It took a young professor named Lydia Sohn to carefully compare papers and to document that the same plots were being reused with different labels (she gave a talk about this at the 2004 Sigma Pi Sigma Congress).

This case smelled bad from the beginning. Most physicists hearing about sonoluminescence-produced fusion were immediately skeptical. It was too far out and seemed to violate Sagan's Razor: Extraordinary Claims Require Extraordinary Proof. My guess is that Science would never have published it, except for the claim of confirmation from other groups, confirmation which we now know did not happen and which Purdue cited has scientific misconduct in their review.

Although I should probably not raise this question in this context, I have to wonder: Do engineers approach research differently than scientists? Do the two professions, while using the same language of experimentation and similar methodologies, place different weights on what passes for verification and confirmation? I know an engineering professor, a very good guy who I will not name, who seems to be only interested in proof of principle: If he gets a device working once, then the project is done. Systematic studies for others - he has moved on to the next topic. Is this an individual trait, or is this behavior widespread among our research engineering cousins? As engineering students get less exposure to basic science (at a time when they really need more science) will the perception of research methodology between science and engineering (continue to?) diverge in the future?

No comments: