An unusual scientific paper

Science is generally self-correcting, but usually not in the same person.

Our astronomer recently noticed a paper on an e-print server that was rather unusual.  The author (call him author A) cited an earlier paper of his, said that’s he’d made a mistake in the calculation, and his previous criticisms of two other groups (call them B and C on the one hand, and D and E on the other) were wrong; and in fact all of them were agreed on the matter at hand.

Now, papers concluding that everyone agrees are not rare, though of course they aren’t those that attract media attention.  Nor are papers pointing out errors in previous work.  The unusual aspect of this one was that the author was pointing out flaws in his own earlier work.

That’s not because no one makes mistakes.  It’s because it’s embarrassing to have gone to the trouble of publishing something, especially a paper criticizing someone else, and later find out you’re wrong.  That means you go to a lot of trouble to check your work, test it against known answers, try it in test cases, use any tricks you can to check for flaws.  (Our tutor is laboriously collecting the tricks he’s developed as a researcher in the hope of passing them on to his students.  The latter are too comfortable, he feels, in their conviction that there will always be an answer in the answer key and for the odd-numbered problems in the back of the book.)  A recurring nightmare for any scientist with a complicated calculation or computer program is that he’s missed something along the line, and it will blow up in his face when someone else points out the simple flaw.

The main defense against errors of this sort is the peer-review system, in which a paper submitted for publication is sent out to referees.  These are experts in the field who should find errors, or at least question the questionable.  But e-prints do not go through referees.  There is some filtration to make sure the obvious paradoxers aren’t given a platform, but researchers with some credibility can put things out quickly.  That’s what an e-print is for: to attract quick comments.

And they do.  Many are the e-prints that attract detailed rebukes from other authors, and not a few are withdrawn afterward.  Many are modified, some heavily.  This is part of the self-correcting aspect of science, much-touted though imperfect.

Well, the original publication of author A did indeed attract a detailed criticism from author D.  The later publication of A quotes in detail from it, to show that none of these criticisms is valid.  That is, although paper A1 was wrong, it wasn’t for any of the reasons D brought forward, which (according to A) range from the inaccurate to the incoherent.  Maybe so.  We haven’t gone into the details of the algebra and the physics to work out just what is happening.  It detracts a little, we think, from the motivation of A that he spends much of his time showing how D was actually wrong.

But here’s an open question: in what other branch of human effort would someone go to such trouble to point out his own earlier mistake?