While publishers are desperately trying to maintain control of intellectual property, authors are expending significant time and energy dismantling current distribution systems. Developers of the e-print archives have overtly opposed the print publications model because they claim publishers don't understand what knowledge is important to researchers. Neither can the industry deliver knowledge on the weekly or even daily basis some researchers now expect. These developers propose eliminating publishers altogether as an intervening party between authors and end-users.

Unlike the transition model controlled by professional publishers, disciplinary specific e-print archives offer an entire body of self-published articles. While such raw archives resolve the problems of breadth of disciplinary knowledge, their raw state makes them no more than vast depositories of print texts that are stored online. Value added filtering functions could turn this kind of archive site into heterotopic spaces that might eventually eliminate the need for publishers altogether.

In the short term, however, academic authors still need credits toward tenure and promotion. Even if citations could be "counted and documented" in a modified raw archive, publishing records still serve as the primary means of judging faculty professional development. Hal Varian, Dean of the School of Information, Management and Systems at UC Berkeley, argues that "the publishing record and quantity of publications [of academic authors] is easier to convey to non-experts than quality of publication" ("The Future of Electronic Journals"). Furthermore, Varian argues that since electronic publication is so inexpensive, "[e]ssentially everything should be published, in the sense of being made available for download. "Ex post" filtering processes (which would occur after posting to the raw archive) will necessarily replace current publish-or-reject publication policies.

Varian describes a typical print-based referee model for electronic ex post publication. He imagines the formation of an editorial board, three members of which would review each submission. But rather than reviewing entire 20 to 30 page articles, editorial board members would evaluate only the one-paragraph abstract and five page summary that are uploaded along with the academic articles. The summaries would follow a prescribed format. Reviewers would rate the submissions on the basis of interest and pertinence to the field on a scale of 1-5, with only about 10% receiving the highest ranking. Authors would be notified of their ratings and could elect to resubmit the piece elsewhere if it did not receive a high rating; once accepted, the author could revise but not withdraw the article.

Varian envisions readers scoring articles as they read and potentially contributing critiques or supplemental narratives and emendations to correspond with the texts. Still other researchers may build hypertextually linked webs that draw upon previous papers to build their own arguments. Or they may develop annotated bibliographies of highly rated articles that would offer networks of rated material for disciplinary study.

Such a model has several advantages over the for-profit transition model. While still sequestered within a limited body of disciplinary knowledge, the non-exclusionary publication policy means that huge volumes of lesser quality research will also be linked to this raw archive. Low rated pieces may not offer professors much in the way of credit toward promotion and tenure, but high rated pieces might be considered as serious contributions.

However, lower rated entries should be included as potentially valuable material--embedded and later cited within the larger archive of works. As concepts reemerge in the work of others, original authors would have opportunities to revise and resubmit arguments. There would not be any doubt as to original "ownership" of ideas because hypertextual links would be established to original versions of documents. But authorship of disciplinary knowledge would become so collaborative that ownership of texts would be difficult to determine. So long as no expectation of capital return was expected for journal submissions, issues of ownership would not be of great concern.

Another advantage of this self-publishing editorial system is that it eliminates the costs associated with current print journal production processes. Editing would occur as a natural part of the review and revision process as the article became important to the larger corpus of disciplinary knowledge. Those articles that remained unrecognized or unused would simply remain in rough form. The only real costs that would be incurred by this system would concern storage and remote access. These kinds of costs could be covered in a host of ways from professional organizations, university or home library support stipends, or perhaps third party vendors.