That is what Bradley Voytek asks in this blog post, from his blog Oscillatory thought. And, in extension (to some degree), what are those static journals for, now that we can discuss things quickly via the net?
He links in to a couple of discussions on Peer Review from 2006 (behind paywall – make sure you are logged in via the university), and goes on to eviscerate some of the arguments.
Clearly, peer review has to some degree become tribal. Not a protector of science, but reputation management, and a place to do turfwars. Well, that’s what happens when a great ape (or, like John Hawkes would insist – hominid) do anything, because status and reputation matters. Of course, it is not the whole bit – peer review does do something more, or at least should help the science be more reliable.
But, it may be time to reform this bit too.
John Wilkins, (who would say, we are so apes), from Evolving Thoughts reiterates this anecdote of editor violations, and there are other anecdotes of the same ilk.
But, they are anecdotes – discussions of egregious violations of the trust bestowed the editor. THere will always be a minor frequency of cheats and non-cooperators, which does not mean that the system is broken.
I also saw, in my twitter feed (no link right now, as I do not know where) a suggestion that statisticians should be involved in peer-review, considering how much crappy statistics are used.
Perhaps we need to realize that the areas of interest now are so complex that doing good work does require teamwork, and a statistician should be on that team – just like (according to Richie Davidson, but I do believe that is the fact here in Lund too) that you keep physicists on hand for your MRI work, so you don’t inadvertently break some of the laws of physics in your interpretations. (Physics itself will not be broken, of course, but your results will be kaput).