The recent announcement of a new journal sponsored by the Howard Hughes Medical Institute, the Max Planck Society, and the Wellcome Trust generated a bit of discussion about the issues in the scientific publishing process it is designed to address—arbitrary editorial decisions, slow and unhelpful peer review, and so on. Left unanswered, however, is a more fundamental question: why do we publish scientific articles in peer-reviewed journals to begin with? What value does the existence of these journals add? In this post, I will argue that cutting journals out of scientific publishing to a large extent would be unconditionally a good thing, and that the only thing keeping this from happening is the absence of a “killer app”.
[emphasis in the original] This reminds of the discussion between Melody Dye and Jason Goldman on bloggingheads.tv. After reading Mr. Pickrell's case, and generally slouching toward a more "open science" stance overall since that diavlog, I think I am now moving toward Melody's position of "end it," rather than Jason's position of "mend it." The action in the Genomes Unzipped post is really in the comments, where Mr. Pickrell' is mixing it up with supporters and detractors. I highly recommend that you check it out. My own thoughts moving from specificity to generality: - Mr. Pickrell notes that despite PLoS's attempts to facilitate comments and interactions, there's a clear "empty restaurant" syndrome at work. It reminds me of Chris Surridge's lament in January of 2007 on the Official PLoS Blog about this problem, contrasting it with the vibrant discussion at Gene Expression Classic. I don't think anyone can deny that PLoS "get's it" on a deep and fundamental level, but over four years on it still seems that the loci of public discussion is distributed across weblogs and not around the papers themselves. - Much of the debate is really only relevant to biological science (ergo, PLoS's biological focus). Physics has arXiv and the social sciences have SSRN. There are many fields where getting the paper in the journal is more of a formality, a stamp of approval which occurs after scholarly consensus has weighed in. Many of the objections to Mr. Pickrell's broader argument might be examined in light of the outcomes in these other sciences which have moved much further along the direction he suggests. - Metrics by their nature are going to be "gamed." I have no doubt of it. The only thing I would offer is that it often takes time for a new equilibrium to be attained, where selfish actors dominate the landscape.