This recent piece from Inside Higher Ed does a good job of describing some of the evolutionary issues I brought up the other day in our research methods class when we were talking about the "conversation of scholars" that takes place via peer-reviewed journals.
To me the most interesting aspect of this is the connection with academic prestige. Note the comment in the opening paragraph of the IHE piece: the move away from print is a "crisis" for reasons that are mostly extrinsic to the publication medium. In other words, the lack of books is not the problem, it's what depends on book sales. So many parallels with the similar "crisis" in the music-recording industry; in both cases, "success" has been defined in terms that are inherent in an industrial process (pressing books and CDs) that is only indirectly related to the quality of the content.
When that industrial process changes, we're left with the Vetting Argument: publishers serve a valuable purpose by filtering content. But what value is that filtering if it emerges partly from a resource-scarcity model? Does Journal X only publish five articles because those are the only 5 good ones submitted, or because that's all that the current issue can support?
I don't know, but it's an interesting question.
Related developments can be seen in the British Medical Journal's decision to "out" reviewers (the article discusses several criticisms of peer review), Nature's trial of "open peer review", and the publish-then-review model used by Philica. The latter is not so outlandish as it seems: if you know that it's not the act of publishing but the reaction of an audience of peers that matters -- and that reaction is completely open for analysis -- it may lead us to a different model, wherein nearly everything (dreck and all) gets published; what would matter most would then be who commented on it and what they said about it. Further, that "reputation" of a scholarly work could continually evolve over time in a very connected and accessible way.
One possible outcome might be the reduction (or even elimination) of the journal's mediating role -- not a welcome prospect to many, I'm sure. But think of it for a moment: publishing via a standard rather than a publishing house, and letting the reviewers make their cases as to the validity of the message. It would certainly be more difficult to identify quality research if we could no longer turn to editorial boards.
Is that entirely bad?