Seed has a new article out by Stephen Ornes, called The Proof is in the Blogging, about the way the story of Penny Smith’s solution to the Navier-Stokes problem played out here. I’m quite fond of the photo included in the Seed article.
I’m not sure there’s anything more to be said about the Navier-Stokes story. One of my colleagues pointed out that mathematics is one of very few subjects in which bringing together a bunch of people with opposite views on what is true generally leads to one or more of them agreeing that they were wrong.
There’s also a short article about this on Slashdot. Taking advantage of the arXiv trackback mechanism, the author found the discussion of this on Lubos’s blog. I was going to take the opportunity to complain about the arXiv censoring links to this blog, but it turns out in this case there is one there. The ways of the arXiv are endlessly mysterious, I have no idea what their trackback policy is these days.
Maybe it’s also relevant to mention that for some reason the hot news retailed here about the proof of finite generation of the canonical ring is not attracting the kind of attention indicated in the Seed picture.
Perhaps it’s only hep-th which doesn’t give you trackbacks?
It is interesting that the recent Navier-Stokes controversy was played out here rather than sci.physics.research or sci.math.research, where it more naturally belonged. I think that this is partly because the moderation takes too long on these newsgroups (some of my posts on SPR have taken days). If they adopted Peter’s policy of allowing posts to appear before deleting them rather than just preventing what they see as irrelevant/obnoxious/uninformed posts at the outset, the whole thing might work better.
Also, the collective quality control exercised by the moderators seems to be less effective than Peter’s personal control. A bit like Communism vs. Capitalism.
I didn’t have much to say during the Navier-Stokes case as I couldn’t think of anything helpful to say. I’m sorry about the way it played out though. My thinking is it wasn’t as bad as it seemed to the people most emotionally involved in it, but I guess it must have felt that way.
Discussion boards take a lot of getting used to. People probably take them more seriously than they should. (Some grow out of this and some do not.)
I honestly don’t see what all the fuss is about. The system worked exactly as it was intended. A preprint was posted on the arXiv, people viewed it, discussed it, found problems, and gave the author quick and concise feedback. The preprint was withdrawn, and the author can begin working on any problems found.
The system works.
Compare the quick turnaround to the tradiational publishing ringmarole. It could have taken Smith over 18 months to find out her paper contained a flaw. The flaw may never have been found at all. And if the paper, or the next version, had no flaws, the journal may have had problems publishing it anyway.
The only problem in this paticular case is that Nature rushed the story to print. If some less than professional scientific publications choose to base their stories on preprints, that could deter people from posting preprints online so as to avoid any publicity and potential embarrassment, and that would be a big setback for the very good thing arXiv has got going.(Aside from unrelated censorship issues)
The system, in the sense of arxiv.org, works. I don’t think any of the fuss is about that.
What stunned me was seeing how too many bloggers, commenters on blogs, and “official” media such as Nature jumped on the bandwagon and touted Penny’s work well before anybody with serious credentials had a chance to do a serious assessment of her claim. People who appear to have no expertise in the subject and have no way of understanding what was in Penny’s paper posted their own personal (always positive) evaluations about Penny’s work.
You might ask what’s the big deal, but some of the people who did this promote themselves as people who know math and how to judge what good math is. So their misleading statements create a higher level of FUD than what we sheltered pure mathematicians are used to.
All in all, I think as math gets more publicity and attention, mathematicians like me are going to have to get used to the extra noise. Overall, I’d rather have the attention. But I think a lot of my colleagues feel differently.
Yesterday, there was a retraction of a paper on math.ph from 2004 on smooth solutions of 3D Navier Stokes.
PS That photo is great. You should have it framed 🙂
This was an extraordinary case. What I find annoying are the many small mistakes present on arXiv. They get corrected in final published versions, but i) the most read version is the first one; ii) too many authors think that fixing errors in revised version without indicating them in the “comments” line is a good idea.
Please, instead of just writing “Comments: 26 pages, 7 figures, LaTeX” add “sorry, when we posted version 1 we were drunk, now version 2 is ok” and I will trust more your future works.
Now that I have read through all the comments on NS it does appear that Penny overhyped things quite a bit. And while the system of peer review did work as intended, the negativity she got is purely of her own making.