ACRL News Issue (B) of College & Research Libraries 794 I C&RL News ■ October 2000 SCHOLARLY C O M M U N IC A TIO N Ivy Anderson Gail M cMil lan & Ann Schaffner, editors Digital archiving Whose responsibility is it? by Deanna Marcum I n the world of paper, preservation has been the clear responsibility of libraries. While not every library has made it a hi priority, many institutions—individually or consortially—have committed resources to preserving the materials they have acquired that have long-term value. However, this model of preserving acquired materials does not apply easily to electronic journals, which publishers typically license to libraries for use. Because so much of the scholarly record is found in journal articles, and because li­ braries are relying more on providing access to the electronic versions than on acquiring the paper copies, there is growing interest in establishing archival repositories for electronic journals. In 1996, the Research Libraries Group (RLG) and the Commission on Preservation and Access issued a report. “Preserving Digi­ tal Information.” based on the work of the Task Force on Archiving Digital Information, co-chaired by D onald Waters and John Garrett. The task force recommended the establishment of certified digital archives— agencies that spell out how they are preserv­ ing digital objects and that can be held ac­ countable for fulfilling their promise. The task force also called for a fail-safe legal mecha­ nism that would allow these certified archives to rescue digital files that are in danger of being lost. gh Perhaps the call for such a formal mecha­ nism was too jarring to the library commu­ nity four years ago for specific steps to be taken. However, as libraries rely increasingly on licenses for electronic journals, both li­ brarians and publishers have become more aware of the potential problems. When ne­ gotiating licenses, librarians are quick to ask what the publisher has done to ensure long­ term access to the material. Since libraries traditionally have financed preservation, pub­ lishers have been forced to consider new business models that include the cost of pre­ serving digital information. W ho is re sp o n sib le fo r a rch iv in g ? Archiving should be accomplished through a set of standards and practices set out jointly by librarians and publishers. To be of lasting value, electronic journals must become part o f th e p e rm a n e n t in te lle c tu a l rec o rd . Archiving efforts must take into consideration all the parties involved—readers, authors, publishers, libraries, and the scholarly com­ munity. To be successful, an agreement must be reached that allows publishers to make a profit, libraries to provide ongoing access, and users to take advantage of intellectual content. In much the same way that standard prac­ tices are in place for print sources, parallel About the author Deanna M arcum is president o f the Council on Library an d Info rm ation Resources, e-mail: dmarcum@clir.org mailto:dmarcum@clir.org C&RL News ■ October 2000 / 795 Editors' introduction changes in scholarly communication and technology have made librarians, publish­ ers, and researchers re-examine their roles in this complex process. Nowhere is this more important than in the area of archiving. In the pre-electronic era, archiving was clearly the role of the library. Long-term pres­ ervation of knowledge was one of our most important professional values and was re­ flected in our practices and organizations. Paper publications were effectively archived in a distributed informal system. Most aca­ demic libraries participated in this system by preserving materials locally. With the advent of electronic publications, the issue has become more complex, and old assumptions no longer necessarily hold. Libraries are often reluctant to take on the task of archiving electronic materials. The level of redundancy needed for archiving in the paper world seems unnecessary in the electronic environment. Even for the largest academic and research libraries, the complexity of the technical knowledge needed is an impediment to tak­ ing action. The prospect of migrating mas­ sive amounts of information on a regular basis is daunting. The difficulties of archiving a more interactive and less static medium are enormous. The temptation is to elimi­ nate all redundancy or to rely on publishers or aggregators to perform these functions. But do these organizations have the same commitment to archiving as libraries did in the past? Are there opportunities for part­ nerships in this new environment? What role can aggregators, national libraries, and li­ brary organizations play? We’ve asked Deanna Marcum of the Council on Library and Information Re­ sources to address the issue of archiving in this column. Marcum is known to most of you for her years of work on important li­ brary and information issues for the Coun­ cil. She has been deeply Involved in recent discussions about archiving electronic pub­ lications. We’re delighted that she has agreed to share of her perceptions with us in this issue.—Ivy Anderson, Gail McMillan, a n d A n n Schaffner mechanisms must be implemented for elec­ tronic journals. Who is working on projects and models? RLG, OCLC, and several international groups are deeply engaged in digital archiving. RLG, OCLC, and the Cedars Project (UK) are ac­ tively exploring metadata standards for digi­ tal objects. An international workshop on the subject is scheduled for late fall. CLIR has commissioned several reports on approaches to digital archiving.’ Individual institutions concerned about both “born digi­ tal” materials and their reformatted digital collections are setting requirements for ar­ chival repositories. Despite all of the good work that these organizations have done to define the prob­ lems and suggest possible solutions, there has been a dearth of digital archival reposi­ tory experience to guide us. CLIR, in collaboration with the Coalition for Networked Information (CNI), convened a meeting of librarians and publishers in the fall of 1999 to answer the question, “What would be required to make electronic jour­ nals accessible for 100 years?” (We are confi­ dent that print materials that have been re­ formatted to microfilm will last a century.) Following the initial meeting, CNI hosted a workshop on the subject for a larger num­ ber of libraries and publishers. The partici­ pants, working from the premise that it is About the editors Ivy Anderson is co o rd inat o r fo r D ig ita l Acquisitions a t H arvard University, e-mail: ivy_anderson@harvard.edu: Gail M cM illan is h ea d o f th e D ig ita l Library a n d Archives (fo rm e rly th e Scholarly Com m unications Project) a t Virginia Tech University, e­ mail: gailmac@vt.edu: A n n Schaffner has been an academic librarian f o r m ore than 20 years a nd is currently a fu ll tim e MBA s tu d e n t a t Simmons College, e-m ail: ann.schaffner@ simmons.edu mailto:ivy_anderson@harvard.edu mailto:gailmac@vt.edu mailto:ann.schaffner@simmons.edu 796 /C&RL News ■ October 2000 necessary to create a Standard for digital re- p ositories, d iscu ssed w hat that Standard should entail. At this workshop, the participants ex- plored technical, practical, and theoretical questions. The group was asked to consider whether a Statement regarding acceptable formats should be released, whether best practices should be defined, and what makes one journal more suitable for archiving than another. Other issues brought to the fore- front included how the repository would function financially and how to establish a minimum level of public access. The discussions were difficult because the stakes are high for all parties. To make progress, we concluded that it was impor­ tant to consider the publishers’ and librar­ ians’ views separately. To provide a point o f departure for the d iscu ssions, Dan Greenstein, director of the Digital Library Federation, and I extracted minimum crite- ria for archival repositories of electronic jour­ nals from the Open Archival Information Sys­ tem (OAIS) reference model and modified them to reflect the specific needs of the li­ brary, publishing, and academic communi­ ties.2 The eight criteria set forth highlight the importance of working together with pub­ lishers to create an environment where the free flow of information will benefit all par­ ties. Agreed-upon minimum criteria are es­ sential to ensure the preservation of intel­ lectual content. Proposed minimum specifications The archival repository must exist as a trusted third party that conforms to minimum re- quirements that both scholarly publishers and libraries agree upon. This will provide a benchmark against which service can be measured, validated, and, above all, trusted by the libraries and publishers that rely upon them. The repository will define its mission with regard to the needs of scholarly publishers and research libraries. Repositories will also be explicit about which publications they are willing to archive. A repository needs to spell out the scope and nature of materials it aims to collect, the strategy and methods it adopts for developing its collections (at- tracting deposits), and the community of li­ braries and other users it seeks to serve. Registries should also be developed that document what scholarly publications are archived and where. Once an agreement is reached, the re­ pository will have sufficient control of de- posited information to ensure its long-term preservation. The repository will negotiate and accept appropriate deposits from schol­ arly publishers. It will negotiate the requi- site perpetual licenses and rights, even as they change over time. The repository will follow documented policies and procedures that ensure the information is preserved against all reasonable contingencies and that enable the information to be disseminated as authenticated copies of the original or as traceable to the original. Because preserva­ tion practices are likely to vary across re­ positories, it may be useful to request that participants in any such coordinated effort agree to document the practices they adopt and make them available for community re- view and evaluation. The repository will make preserved in­ formation available to libraries. Publishers seem willing to deposit electronic content in archival repositories only under certain conditions. Although repositories will, need to support access at some level, they should not replace the normal operating services through which digital scholarly publications are typically made accessible to end users. The repository will ensure that data can be disseminated to libraries in a renderable form. At a minimum, libraries should be able to create end-user services appropriate to the disseminated data and to do so inde- pendently of any assistance from those who initially produced the data. Repositories will work as part of a net- work. Libraries may benefit from common finding aids, access mechanisms, and regis- try services that are supported by a network and allow uniform identification and access to information. Publishers may benefit from having access to a single repository or group of repositories that specialize in publications of a particular type and the cost effective- ness that results from such a network. Technical requirem ents Without an established minimum criterion for preserving electronic journals, multiple C&RL N ews ■ O cto b e r 2000 / 7 97 overlapping versions in fragile, rapidly chang- ing environments are at risk o f loss. The solu- tion is to create a system that allows for tech- nical growth. Everyone agrees that redundancy is essen­ tial. By requiring three separate, independent facilities to act as repositories, the documents can be protected from human error and natu­ ral disaster. If at least two o f these repositories exist in different countries, governmental poli- cies that may interfere with content are no longer a concem. Regardless of the techno- logical implementation chosen, the materials must exist in a safe haven where they can be accessed over time. How access should change or not change in the networked environment is central to the issue of technical development. Once Standards are set, a number o f events will occur. Pub­ lishers can make business decisions about the electronic format. Libraries will be forced to evaluate the archivability o f electronic joumals when making decisions about purchasing and collection development. Authors will have better information for establishing expectations about the long-term availability o f potential Publishing venues for their works. Scholarly societies designing electronic Publishing pro­ grams will have guidelines for addressing ques- tions o f archivability of their publications. Once disseminated, these copies can po- tentially be maintained in perpetuity without further permission from or even interaction with the publisher. One major caveat is the need to ensure a limited level of permanent public access to these copies by the broad scholarly community while the work remains under Copyright. Protocols must be set to rec- reate these mechanisms for electronic journals. Once Copyright has expired, the materials must remain accessible. How can individual librarians m ake a d ifferen ce? With so much o f the scholarly record now being distributed in electronic form, we risk losing an entire generation o f accumulated knowledge. If no one is certain whose respon- sibility it is to preserve electronic joumals, no one will be responsible. Publishers recognize that librarians are concemed about this issue, and they are, happily, beginning to issue their policies about digital archiving. (continued on page 807)