There has been a lot of discussion in the science bloggysphere about this post by Joe Pickrell (and see his related post for background) which discusses possible ways to improve scientific (that is, technical, not popular) communication. It’s interesting, but it seems separated from what scientists use papers for (as currently construed).
I’m not convinced that the de facto purpose of publication is (note: is, not should be) is to communicate results to other scientists. Realistically, when someone publishes a paper, is she thinking, “Now all my colleagues can hear about my research!”? Or is she really thinking, “I’m that much closer to getting my R01 grant renewed”? Obviously, papers are useful, especially for older projects (or those belonging to dead scientists…). But, if you’re like me, you probably have a stack of unread papers, or a folder of pdfs (or both) that you never get around to reading. It’s not a very good system to disseminating information when you get right down to it.
But where publication, as it exists today, is critical is in the funding process. Most funding agencies will only view ‘legitimate citations’, both in terms of justifying your work as well as establishing productivity, as those papers that are published in peer review journals (until recently, even papers in press didn’t count–they had to be published). If we want to move towards a better publishing model–and I would argue that should include considering new formats–then we need to get funders on board with various proposals, or else those proposals will not get very far.