I tagged Steinn's post on publishing a comment a few days ago, because I thought it was pretty funny. In the interim, it's been picked up by the usual suspects as more evidence of the need to completely discard the current publishing model in favor of something more blog-like.
None of the subsequent discussion has answered what, to me, seems like the most obvious problem with the original story. Namely, why the insistence on publishing this as a Comment in the first place? I mean, here's the start of the saga:
1. Read a paper in the most prestigious journal in your field that "proves" that your entire life's work is wrong."
2. Realize that the paper is completely wong, its conclusions based entirely on several misconceptions. It also claims that an approach you showed to be fundamentally impossible is preferrably to one that you pioneered in its place and that actually works. And among other errors, it also includes a serious miscalculation - a number wrong by a factor of about 1000 - a fact that's obvious from a a glance at the paper's main figure.
To me, this doesn't sound like something that demands a short Comment in reply. This sounds like an excuse to write a full-length debunking paper in "the most prestigious journal in your field." Cha-ching!
I mean, obviously the subfield, whatever it was, is important enough for a post on the subject to make it into TMPJIYF, so that's a factor on your side. Your response paper would be directly contradicting something previously published in TMPJIYF, which is another clear indication of the importance of the work. If it's "your life's work," you ought to have plenty of supporting material to back up your version of things, and possibly even extending the basic idea in a way that would make the paper even more attractive to TMPJIYF.
And given the 120-odd steps following those two, it clearly wouldn't be any more work to publish the thing as a full-blown paper rather than a Comment, which would avoid the fifty-odd steps concerned with the length limit for Comments.
Plus, it would look better on the CV than a Comment. And probably would be read by more people-- I have a tendency to forget that the Phys. Rev. journals even publish comments in the first place, and it never occurs to me to scroll all the way to the bottom to see what comments have been published in the latest PRL.
It may be that there's some discipline-specific reason why this would be totally inappropriate-- they're clearly not working in my corner of the physical sciences, in which the most prestigious journals limit full papers to around four pages, basically the same length that the original author wants Comments to be. But, really, I don't see why, other than sheer bloody-minded stubbornness, this should be a Comment rather than a full paper.
I think this is a case of "at any given step it is easier to publish/finish/push through the Comment" - rather than go back and write a full paper - also a paper reaffirming an existing result and pointing out a false result can easily run into a referee trap of not being original/relevant and be bumped down the journals.
It is a flaw in the system, in that they were not claiming anything new, just saying "we were too right".
Congratulations, this page is now the only google result for "TMPJIYF."
Wouldn't you need some kind of new data in order to publish a new paper? If the refuting argument is based on data that's already been published (and on the new paper screwing everything up), it seems like a Comment/Reply is the way to go.
(I've seen both done, together, actually. A full paper with new data, plus a cranky comment-and-reply, all in the same journal. But maybe the editors were sympathetic to the commenters in that case - the paper that I'm thinking of seemed awfully hasty, and went through review surprisingly fast for my field.)
But that takes way too much time...
If you publish it as a comment, at least in Phys Rev online, the comment appears right under the article abstract. A reader will see "there is a comment on this paper" without needing to click through the "cited-by" link and scroll through a whole list of things.
In other words, if you publish a separate rebuttal paper, you'll still have people who do some lit search, come across the original (wrong) article and take it at face value. (Not people in the specialty, but people on the periphery.)
If you publish as a comment, browsers hit the original paper and *your* rebuttal---and not the swarm of other citations, if any---at the same time.
At least under standard scientific publishing norms, it's hard to publish a full paper that simply restates something that has been well known for yonks and is already established in the literature. Any reviewer worth their salt will say "this is well known". It is only by tying the paper to the incorrect one that the publication is justified.
In principle, of course, a comment should be much easier and quicker to write, cos you only have to point out a significant error, rather than make a substantial contribution to science. If journal editors would just do their job properly, it would work fine.
#6 hits it I think.
Comments are (or really really should be) attached to the original publication in an easy to access manner.
Most of us read articles online anyway, so doing this well should be trivial.
Having peer-reviewed comments is valuable. Having comments which are just moderated is also good, but they are certainly different. Maybe the former could be "Replies" and the latter "Comments".
We are also hitting upon my current biggest annoyance with publishing... dumb submission hurdles. Everything should be much easier to submit. Physicists have it relatively good; try dealing with the primitive dead-wood publishing centric (or seemingly random arbitrary) rules Bio journals impose.
I get that journals should impose rules for uniform formatting and style, but we've have very nice ways of dealing with those for a good long while... None of which involve Microsoft Word (LaTeX and BibTeX). Online-only publishing should be even easier.
PS: Which reminds me, I've got a couple of papers to review ;)
Just to clarify. IMO: LaTeX is good, Word is bad.
I just discovered that my co-author wants to submit the paper I've been writing to Word centric publication. I don't even own a copy of Word! I'm seriously tempted to submit it in RTF.