Columbia Working Group on Electronic Texts

Convenors: E. Sloan and A. Okerson
28 September 1994

Approximately 35 representatives of universities, university libraries, university presses, and learned societies met at Columbia University on 22-23 September 1994 to discuss the future of campus-based electronic publishing. This document represents a consensus of those discussions and is presented as a talking paper for further action.


The incentive to make substantive change in the way the scientific and academic publishing system works is felt at present most strongly in the university communities by librarians and by university presses. This sense of urgency defined and helped bring together this particular collection of people. Academic authors and readers do not yet share the urgency for change, not least because librarians have done their jobs too well: careful trimming of acquisitions at the margin and expeditious systems of ILL and document delivery have so far shielded most academics from the present crisis, still more from the upheavals of the coming transformation of information into electronic form.

The task of this group, therefore, is to find ways in which the parties who already have the incentive can take concrete actions to begin to change the culture. Libraries and university presses cannot create or mandate a revolution, but they can think carefully about how their contributions can affect the whole system. If a critical mass is needed in order to see real revolution, the first question is how best to create, or encourage the creation of, individual molecules suitable for that reaction. What we hope to make possible is the creation of some new alliances among participants in the scholarly publishing environment with compatible and complementary interests, and to give those alliances strategic vision and some tactical guidance. If we can be more deliberate and less driven, we may make faster progress.

The traditional stages and sequence of preparation of scholarly publications seem to us to be five in number:

  1. creation (notionally an author's function)

  2. organization (peer review, adding to the publicly-indexed stream of documented material -- a publisher's function)

  3. preparation (a publisher's function)

  4. storage (a librarian's function)

  5. cataloguing/indexing/abstracting (a function that may be performed by publishers, librarians, or independent agents)

  6. use (a scholar's/scientist's function), but with support of user training (a librarian's function)

But we believe that stages 3 and 4 in this sequence can often now be brought in-house to the home institution of the scholar, with real advantages in economy and in the timely and useful flow of information, while stage 2 becomes the fourth stage in a revised sequence. Among other things, this means that "use" can be made of such materials in appropriate ways at earlier and timlier stages in the process. That is, material for teaching or material for informal discussion can be more widely and readily circulated while the peer review and formal publication processes continue. This insight builds on and seeks to strengthen and accelerate the vision that underlay traditional paper preprints and that has animated the exciting experiments in e-preprints that are already gaining attention in specific fields.


This listing is not comprehensive or exclusive, but represents a distillation of some of the common positions held by participants at this meeting and is included as a way of suggesting the long- term objectives that our proposals intend to advance towards.

There is need for more kinds of information to be published quickly, easily, cheaply (e.g, conference proceedings).

We need to combine scholarly information with scholarly discourse through commentaries, moderated discussions.

Primary source material can be used more widely and easily if it is in digital form.

Electronic scholarly information should be easier to access and use than it is now.

Electronic scholarly publishing should count toward tenure.

Permissions and use of material must be facilitated.

Model site licenses and contracts should be designed to make materials readily available for all kinds of use by scholars.

Distribution to 3rd world countries should be facilitated.

How can the economic model be shifted to: save money in production and distribution, to fund and distribute review publications, to continue to fund the editorial review process?


1. Individual institutions will provide electronic publishing services to their faculties, to include: --assistance in "manuscript preparation" up to and including software support and tagging in SGML or other appropriate markup language; digitizing of images; information and assistance about standard formats for networked distribution of information; --guidance in locating or creating electronic repositories of texts, including "preprint servers", "electronic journals", and "electronic text centers" appropriate to discipline and subject; --space and support on a server to make published material directly available on the Internet for their own faculty from whatever discipline; --space and support on a server for materials from whatever source (on or off campus) in specific disciplines or sub- fields; this would involve development of local bibliographical expertise in these areas, but would have the advantage of making that expertise available beyond the single campus, and of providing support of the kind recommended here indirectly for independent scholars, scholars at smaller institutions, etc. -- costs to such participants would thus be lower than those of attempting to acquire and maintain such services on their own; --assistance in getting copyright clearance to use materials, as well as advice and encouragement on the possibilities of fair use.

2. These services should be supported by a revised and reconsidered institutional policy regarding Intellectual Property. Faculty who use the common e-publishing service could be required, e.g., to submit for publishing elsewhere only on condition that certain rights be retained at the home campus (e.g., that of unlimited distribution of the e-form of the publication to members of the home community); faculty who retain that prepared material on university-owned or -supported servers including the common server of last resort could be required to consent to a still more liberal policy of distribution. Faculty who choose not to avail themselves of the services would retain more conventional rights to sell their work to the highest bidder on whatever terms that bidder sets. The follow-on AAU/ARL task force for Intellectual Property could write model policies for use in these several settings, as well as model contracts for authors participating in projects under this egis.

3. Texts suitable for preparation in this way would include but not be limited to: --materials individual faculty wanted to use in teaching; --materials ready for pre-preprint circulation, that is to say, rough and venturesome ideas not ready for prime time but seeking discussion; --materials closer to traditional preprint status, that is to say, ready for publication but not yet peer-reviewed; --published materials, that is to say, peer-reviewed. --postprints, that is to say, material published previously and now made available in electronic form, perhaps with revisions as a virtual "second edition", for wider distribution.

4. The several grades of material could be flagged by the authors with certain implications. Materials for teaching only could be made available at the preparer's discretion on a restricted basis on campus for restricted periods of time. Pre- preprint materials would be available for a limited time at the author's discretion to discussion lists, usenet groups, and the like, but not archived. Once the "preprint" flag is placed on material, however, it could be required to be archived and made permanently available (institutional policies will have to think about this, and post-publication preprints could be tagged with reference to the definitive publication); the advantage to the scholar of using this flag is that it signals material of greater interest to scholarly searchers/indexers elsewhere. Finally, the "published" flag would only be added when someone other than the author adds it to a recognized database or publication, at which point the "publisher" would take primary responsibility for archiving that form of the document.

5. In may respects, the services that we recommend institutions provide can be done by judicious organization of existing resources. We naturally think of the library as the logical location for such service, but local cultures vary, and on some campuses, the natural home may lie in institutional "academic computing" or even distributed among the computing arms of various individual schools and colleges within the larger institution. Existing sources of support that can and should be tapped for such activity include but are not limited to: --user services (provided by academic computing, library, etc.) --faculty development funds (from academic deans: including funds for creating new courses and course materials, for supporting faculty research, esp. for supporting junior faculty work towards promotion, as well as funds targeted for helping older faculty reignite stalled careers) --instructional support funds, especially those designed to equip and make optimal use of electronically enhanced classrooms --library electronic text centers

6. Putting the material thus created into wider play will require activity beyond the level of the local institution and the local library. We suggest that collaborations with learned societies offer the most fruitful way to pursue this goal from within the universities. Some disciplines will require very little such support (it was suggested at our meeting that the support provided its members by the American Math Society, for example, largely obviates need for them), while many humanities disciplines require urgent support. (The cooperation between Emory, represented at our meeting, and Scholars Press looks like one fruitful place to pursue this collaboration). The roles of societies here could include: --peer reviewing and publishing electronic journals in sub- fields more specialized than their flagship journals can support; --designing conferences that draw upon e-preprints for more informed and advanced discussion of current work; --educating scholars by example about the advantages of this kind of publication; --designing and advertising enlightened and relatively generous IP policies, encouraging the freest possible flow of discourse among scholars, and thus setting a model with which commercial publishers will be forced to compete.

7. None of these services will be of value unless scholars take advantage of them. At the local institutional level, attention must be given to incentives.


1. The natural environment for this initiative is in the follow-up to the AAU/ARL task force reports. The support of key AAU leaders can make these projects happen on individual campuses and can facilitate the kind of cooperation to ensure compatibility of projects, comparability of Intellectual Property policies, etc., that will much assist success. The other cultural change most urgently needed is that which supports recognition of innovative electronic publication within the academic reward system, and this can best come from recognized academic leadership.

2. With AAU support, it is reasonable to think that we could move over a year or two to the point of getting significant outside funding to support some ambitious initiatives. The Columbia group should develop this document as a concept paper to guide the AAU in that direction, and should remain active as fosterers, encouragers, kibbitzers in this area.

3. At the same time, there is no need to wait or delay. Much of what is outlined above can be done, at least on a modest scale, with existing resources. What we might call "pilot pilot projects" can and should be started up as soon as possible, to test some of the ideas here, to show what can be done, and to begin seeding incentives and ideas among faculty authors and readers. Members of the Columbia group should take particular responsibility in this area, but those who expressed interest in attending but were unable to do so should be brought into the stream as soon as possible. In particular, a project team distilled from this group can serve to facilitate communication among projects and to do some limited technical coordination to attempt to assure maximum interoperability among congruent enterprises. (It would also be a strong point of such projects if one or more of them contained significant multimedia components.)

4. University presses already have a considerable investment in expertise and staff. Discussions should begin now at the institutional level, starting with members of the expanded Columbia group, between libraries, presses, and academic computing centers, about how best to apportion responsibility, capitalize on strength, and reduce redundancies.

4. This would also be a good time to approach learned societies to invite cooperation. This can be done formally by approach to the ACLS (perhaps AAU-sponsored at an upcoming ACLS meeting, perhaps in association with the Getty-sponsored initiatives with ACLS for electronic enhancement of scholarship), but also heuristically where contacts already exist, as with those societies that attended the Columbia meeting or other smaller societies with which institutions or individuals in our number have influential contacts. Numerous smaller societies, in particular, are looking with anxiety on developments in this field and would probably welcome constructive collaboration with one or more major research libraries.


What we outline above can be undertaken largely from within the existing base of interest and commitment from which this meeting arises. In order for scientific and academic discourse to reach its audiences effectively, further steps must be taken. We outline here two paths that could usefully be followed by one or more institutions and societies. We separate the initiatives because we felt that some of what we propose already assorts well with activities in libraries and on campuses, while some more closely resembles traditional publisher's roles. What we outlined above builds on what we do now; what we suggest here would more venturesomely take some of the explicit publisher's roles for players on the campus.

Academic server model
This model, proposed on a number of listservs this past summer, recommends that universities obtain newly prepared research papers from their faculty and make them widely available on the Internet. The papers could still be submitted for publication. In fact, papers would be flagged as to their review status: e.g., "preliminary", "under review", "accepted for publication", etc. The ultimate goal, however, would be for the system to overtake and replace the current system of print publishing. At our meeting, there was practical discussion, for example, of a proposal for an online archive at Indiana University containing African Studies material on government and political science. Issues that would need to be addressed in the design of this system include:

  1. attracting the support of a large number of universities so that a critical mass of publications would be available and the practice would be viewed as an accepted one;

  2. providing a critical mass of papers in any given discipline so that researchers turn to the server system routinely;

  3. providing an indexing system that enables scholars to use the system easily;

  4. making contribution to the system sufficiently prestigious so that researchers will want to contribute and placement on the server will not inhibit a paper from being published or receiving credit toward tenure. Thus major universities would need to participate, and would need to engage the contribution of senior faculty;

  5. providing authentication, version control and "labeling", so that quality control and authenticated input can be guaranteed to authors, consistent citations can be guaranteed to users, and review status of papers can be determined;

  6. developing a process to begin to add reviewing and refereeing to the system;

  7. assigning archival responsibility.

"Prestigious" e-publishing model
This model would be based in scholarly societies and universities presses--ideally a collaboration of the two. It would bring electronic scholarly publishing into the mainstream, and would test whether savings can be gained even in fully edited and reviewed publications. The American Physical Society is revamping its publications program in a manner that can serve as a model to others, but smaller, less affluent or less technically sophisticated societies need additional support. A model project was proposed by APSA and the Columbia University Press: International Affairs Scholarship Online. Elements of the model proposal include:

  1. High editorial standards;

  2. Wide range of formats included in content: newsletters, proceedings, articles, book-length pieces, commentary, etc.;

  3. Opportunity for discourse, e.g., a moderated discussion group;

  4. Site licenses to campuses and other organizations that allow users full permission to read print, download, etc.;

  5. WWW-based presentation, so that material is easily available to Web clients;

  6. inclusion of publications of scholarly institutes, etc., providing these organizations with an economical means of distributing information they have been printing. (Fees for this distribution could help support costs.);

  7. Good indexing and ease of use.

It would be important to explore the economics of this model over time. Without the usual income from individual subscribers to scholarly journals, site licenses will not be inexpensive. But the intent is, over time, to replace future more costly paper publications that would otherwise be purchased by those sites.