Let’s separate dissemination from selection

I suspect that scientific publications originated primarily as a means for disseminating research results. Today we live in a bigger and more complex, and perhaps also more number-dominated world, and publications serve also the goal of ranking researchers according to productivity, impact or other metrics [1, 2]. We also live in a highly-connected world, via the internet and the web. I believe that we can leverage this technology in ways that, as far as I know, we have not yet seriously pursued, in order to improve research dissemination. In the long run, these changes could also result in improving review, selection and scientific debate. Dissemination, review and selection are distinct functions of the publication process. The goal of dissemination is to spread research ideas and results throughout the community. The goal of peer review is to check those ideas, in terms of correctness, originality, or other metrics, and provide feedback to the authors so that they can improve their work.1 The goal of selection is to select those ideas and results that seem more worthwhile. Conferences and journals today perform all three functions. Dissemination and selection are usually not separated in the publication process: a paper that does not get selected by the program committee of a conference will not be included in the conference proceedings, therefore will not be disseminated (at least not through this channel). I suspect that the reason for not separating selection and dissemination is historical and pragmatic: in the pre-internet days, paper proceedings or journals were the only means of publication. Only a limited number of articles can be included in a proceedings book or a journal issue, therefore a selection has to be made. Naturally, the articles that are not selected, are not published, and not disseminated either. Today, with electronic publications, this resource limitation no longer exists: the on-line (or even CD-ROM) proceedings or journal can include a virtually unbounded number of articles. Of course, this does not mean that review and selection are no longer necessary. Peer review is a pillar of scholarly work. Selection is also necessary for many different reasons. One reason is that, in conferences, there is still the limitation of time: in a meeting that lasts three days, only a limited number of works can be presented. Therefore, some selection needs to be made.