Stevan Harnad
          Department of Electronics and Computer Science
          University of Southampton
          Highfield, Southampton

An endstate will be described for the world's refereed research literature. It will be argued that this endstate is optimal for both research and researchers, that it is inevitable, and that its coming can be hastened. A practical transition scenario for passing from the current state to this endstate will be proposed. The endstate is a pair of parallel archives in each of which all authors deposit all their unrefereed preprints as well as their refereed reprints: a global, multidisciplinary archive, mirrored and distributed (with a protective distributed coding if needed), plus local, author-institution-based distributed archives. The Global Archive should be modeled on the Los Alamos Archive (indeed it would be most efficient and economical if Los Alamos itself were scaled up to become that Archive). The local archives should be institution-based and interoperable, both with one another and the Global Archive. The world research community will access the literature via the archives for free, just as the authors provide them for free. Journals will have to phase out the costs and cost-recovery mechanism of paper publication (Subscription/Site-License/Pay-Per-View S/SL/PPV) and scale down to the much reduced cost of Online-Only publication, whose only remaining costs will be those of quality control (peer review and editing). These will be recovered out of author page-charges, paid not from authors' pockets but from less than one third of the annual saving by all authors' institutions from the termination of S/SL/PPV. Institutional libraries should first pool resources from S/SL/PPV savings during the transition to continue to cover the expenses of those journals that commit themselves to scaling down to Online-Only, and once the transition is well underway, those savings should be rechanneled to institutional author publication funds, so that the page charges can be paid directly by the authors. The Archive will have overlay sectors controlled by the journals, which will officially authenticate their contents. The multidisciplinary literature will be citation cross-linked for navigation and collectively uploaded and upgraded in perpetuo in pace with the evolution of the technology. Researchers, research, and hence all mankind should be the beneficiaries of freeing the research literature at last from the physical and financial access barriers of the papyrocentric age and launching it into the PostGutenberg Galaxy as Scholarly Skywriting.