Margaret Hedstrom is Associate Prosfessor in the School of Information and Libary Studies at the University of Michigan

Digital preservation: a time bomb for Digital Libraries

Margaret Hedstrom

The challenges of digital preservation

The purpose of preservation is to ensure protection of information of enduring value for access by present and future generations (Conway, 1990: 206). Libraries and archives have served as the central institutional focus for preservation, and both types of institutions include preservation as one of their core functions. In recent decades, many major libraries and archives have established formal preservation programs for traditional materials which include regular allocation of resources for preservation, preventive measures to arrest deterioration of materials, remedial measures to restore the usability of selected materials, and the incorporation of preservation needs and requirements into overall program planning.

Preservationists within the library and archival community have been instrumental in developing an array of tools and methodologies to reduce the decay of traditional materials and to restore books and documents that have deteriorated to such an extent that their longevity and usability are threatened. Provisions for fire protection and adequate environmental controls frequently are incorporated into new library and archival facilities. Rehousing of acid-based paper materials is a common task in many repositories and microfilming is used extensively and cost effectively to preserve endangered materials. Undertakings such as the brittle books initiative, the American Newspapers Project, and the NEH-funded microfilming program have saved millions of unique and imperiled items (Preserving the Intellectual Heritage). Many libraries and archives have curbed their voracious appetites for acquisition and collecting in an effort to balance the breadth and depth of their holdings against long-term stewardship responsibilities. The change over from acid to alkaline paper in publishing and much desktop printing counts as a significant victory for preservation.

Much remains to be done to preserve cultural, intellectual, and scholarly resources in traditional formats that form the foundation for humanities research and teaching. An estimated 80 million embrittled books reside in American libraries, 10 million of which are unique; and countless journals, newspapers, photographs, and documents require preservation treatment to survive well into the next century. Thousands of repositories lack the means for disaster prevention or adequate environmental controls to avoid catastrophic loss of their holdings. The success stories and regular use of established preservation methods are found almost exclusively in developed countries, and within developed countries in preservation of print materials in major institutions (Preservation of Archival Materials).

Digital preservation raises challenges of a fundamentally different nature which are added to the problems of preserving traditional format materials. By digital preservation, I mean the planning, resource allocation, and application of preservation methods and technologies necessary to ensure that digital information of continuing value remains accessible and usable. I intentionally use the term "continuing" rather than "permanent" value to avoid both the absolutism and the idealism that the term "permanent" implies (O'Toole). My concept of digital preservation encompasses material that begins its life in digital form as well as material that is converted from traditional to digital formats.

Recording media for digital materials are vulnerable to deterioration and catastrophic loss, and even under ideal conditions they are short lived relative to traditional format materials. Although archivists have been battling acid-based papers, thermo-fax, nitrate film, and other fragile media for decades, the threat posed by magnetic and optical media is qualitatively different. They are the first reusable media and they can deteriorate rapidly, making the time frame for decisions and actions to prevent loss is a matter of years, not decades.

More insidious and challenging than media deterioration is the problem of obsolescence in retrieval and playback technologies. Innovation in the computer hardware, storage, and software industries continues at a rapid pace, usually yeilding greater storage and processing capacities at lower cost. Devices, processes, and software for recording and storing information are being replaced with new products and methods on a regular three- to five-year cycle, driven primarily by market forces. Records created in digital form in the first instance and those converted retrospectively from paper or microfilm to digital form are equally vulnerable to technological obsolescence.

Another challenge is the absence of established standards, protocols, and proven methods for preserving digital information. With few exceptions, digital library research has focussed on architectures and systems for information organization and retrieval, presentation and visualization, and administration of intellectual property rights (Levy and Marshall). The critical role of digital libraries and archives in ensuring the future accessibility of information with enduring value has taken a back seat to enhancing access to current and actively used materials. As a consequence, digital preservation remains largely experimental and replete with the risks associated with untested methods; and digital preservation requirements have not been factored into the architecture, resource allocation, or planning for digital libraries.

My assigned topic "mass storage and long-term preservation" offers fertile ground for discussing conceptual and methodological challenges facing digital libraries and archives. The two terms "mass storage" and "long-term preservation" embrace a contradiction in the current state of affairs of digital library development, representing a time bomb that threatens the long-term viability of this new type of library. New technologies for mass storage of digital information abound, yet the technologies and methods for long-term preservation of the vast and growing store of digital information lag far behind. Strategies, methods, and technologies for long-term preservation of digital information that already exist or are being discussed, designed, and developed today have yet to demonstrate the technological or economic feasibility of operating on a mass scale. Our ability to create, amass and store digital materials far exceeds our current capacity to preserve even that small amount with continuing value.

Digital Preservation Requirements

In order to preserve digital materials on a scale commensurate with mass storage capabilities and in formats that are accessible and usable, it is necessary to articulate some basic requirements. There are two ways to examine digital preservation requirements: from the perspective of users of digital materials and from the view of libraries, archives, and other custodians who assume responsibility for their maintenance, preservation, and distribution. Libraries and archives will not accomplish their preservation missions if they do not satisfy the requirements of their users by preserving materials in formats that enable the types of analyses that users wish to perform. At the same time, libraries and archives are unlikely to be able to satisfy all requirements of all potential users primarily due to resource constraints. By making explicit preservation requirements from both the users' and custodians' perspectives, libraries and archives will be better able to integrate digital preservation into overall planning and resource allocation.

The potential uses of digital materials are varied, unpredictable, and almost endless (Gould). Any generalizations, even if restricted to one community of users such as humanities scholars, run the risk of overlooking and understating potential user needs. Precise requirements for presentation and analytical tools vary among disciplines, yet some basic requirements are likely to transcend fields and disciplines. The ability to establish the authenticity and integrity of a source is critical to users, whether it is generated by an individual, created in the conduct of institutional business, or produced through a formal publication process (Lynch). Mechanisms that will enable users to establish authenticity require archives and libraries to store much more than the content of digital documents. Attributes such as formal document structures, metadata that document the maintenance and use history of the document, time and date stamps, and a series of references among documents are essential for determining authenticity and for understanding the provenance of sources and placing them in a larger context (Graham).

Michelson and Rothenberg (1992) argue that networking and access to digital sources will change all dimensions the scholarly work process, including identifying sources, communicating with colleagues, interpreting and analyzing data, disseminating research findings, and teaching. If their projections are correct, digital preservation programs must support a high degree of integration of source material into analytical processes by coupling research sources with the tools necessary to analyze them; by maintaining linkages between research results and the sources on which they are based; and by providing a means to incorporate primary sources into teaching. Users will seek documents that are easily retrieved and manipulated, transmittable, and transportable from a repository to the sites of research, presentation, and teaching. It seems safe to assume that humanities scholars will need the capability to search through and select relevant sources from large bodies of heterogenous materials, to compare sources to each other, and to view specific documents at high levels of granularity. Digital preservation will add little value to the research process if it serves only as an alternative form of storage from which analog replicas are produced for use with conventional analytical methods. Preserving digital materials in formats that are reliable and usable, however, will require long-term maintenance of structural characteristics, descriptive metadata, and display, computational, and analytical capabilities that are very demanding of both mass storage and software for retrieval and interpretation.

Digital preservation requirements may be expressed differently by archives, libraries, and other types of repositories that are struggling to meet escalating user expectations with limited financial and technical resources. Storage systems should be capable of handling digital information in a wide variety of formats, including text, data, graphics, video, and sound. Digital storage is not only an alternative means for storing print formats because many types of digital objects do not have print equivalents and cannot be preserved in non-digital formats. Ideally, storage media will have a long life expectancy, a high degree of disaster resistance, sufficient durability to withstand regular use, and very large storage capacities. Conversion from analog to digital formats and migration to new generations of technology will be rapid, accurate, and inexpensive enough to permit very large scale transfers of heterogeneous materials. Storage space requirements will be minimal and not demand highly sensitive environmental controls. To make digital preservation affordable to the widest possible range of organizations and individuals, equipment, media, and maintenance costs must be modest.

Current Preservation Strategies and Their Limitations

Most librarians and archivists have accepted the basic wisdom -- for now at least -- that digital preservation depends upon copying, not on the survival of the physical media (Lesk). But copying, also referred to as "refreshing" or "migration" is more complex than simply transferring a stream of bits from old to new media or from one generation of systems to the next. Complex and expensive transformations of digital objects often are necessary to preserve digital materials so that they remain authentic representations of the original versions and useful sources for analysis and research (Task Force on Archiving of Digital Information).

Current methods for preserving digital materials do not fully support achieving these objectives. When faced with the responsibility for preserving digital materials, archives and libraries face a series of complex and difficult choices based on the format of the original materials, the anticipated uses for it, and the technical and financial resources available to invest in preservation initiatives. I will review some current preservation strategies beginning with the most elementary and established methods and ending with proposals that have not yet been tested.

Probably the most commonly used preservation strategy is to transfer digital information from less stable magnetic and optical media by printing page images on paper or microfilm. It seems ironic that just as libraries and archives are discovering digital conversion as a cost-effective preservation method for certain deteriorating materials, much information that begins its life in electronic form is printed on paper or microfilm for safe, secure long-term storage. Yet, high-quality acid neutral paper can last a century or longer while archival quality microfilm is projected to last 300 years or more. Paper and microfilm have the additional advantage of requiring no special hardware or software for retrieval or viewing. Perhaps this explains why in many digital conversion projects, the digital images serve as a complement to rather than a replacement for the original hard copy materials (Conway, 1994).

Another strategy for digital preservation is to preserve digital information in the simplest possible digital formats in order to minimize the requirements for sophisticated retrieval software. Digital information can be transferred across successive generations of technology in a "software-independent" format as ASCII text files or as flat files with simple, uniform structures. Several data archives hold large collections of numerical data that were captured on punch cards in the 1950s or 1960s, migrated to two or three different magnetic tape formats, and now reside on optical media. As new media and storage formats were introduced, the data were migrated without any significant change in their logical structure. This approach has the distinct advantage of being universal and easy to implement. It is a cost-effective strategy for preserving digital information in those cases where retaining the content is paramount, but display, indexing, and computational characteristics are not critical. As long as the preservation community lacks more robust and cost-effective migration strategies, printing to paper or film and preserving flat files will remain the methods of last resort for many institutions and for certain formats of digital information.

Libraries and archives with large, complex, and diverse collections of digital materials are only beginning to test strategies that normalize various types of holdings by converting digital records from the great multiplicity of formats into a smaller, more manageable number of standard formats (University of the State of New York). A repository might accept textual documents only in one or a few commonly available commercial word processing formats or require that documents conform to standards like SGML (ISO 8879). Databases might be stored in one or a few common formats or converted to a SQL (Structured Query Language) compliant format, while image files might conform to the tagged image file format (TIFF) with standard compression algorithms.

This approach has the advantage of preserving more of the display, dissemination, and computational characteristics of the original materials, while reducing the large variety of customized transformations that would otherwise be necessary to migrate material to future generations of technology. The strategy rests on the assumption that software products which are either compliant with widely adopted standards or are widely dispersed in the marketplace are less volatile than the software market as a whole. Most common commercial products today provide utilities for backward compatibility and for swapping documents, databases, and more complex objects between software systems. Although this strategy simplifies migration and may lower digital preservation costs by reducing the amount of customized reformatting needed as technology changes, it does not eliminate the need for regular migration of digital materials. Software and standards both continue to evolve and even repositories with structurally homogeneous holdings can expect to be required to migrate their digital materials periodically.

Current methods fall far short of what is required to preserve digital materials. All current preservation methods involve trade- offs between what is desirable from the standpoint of functionality, dependability, and cost and what is possible and affordable with current technologies and methods. Consequently, most repositories are coping by employing interim and less than desirable strategies, if they are addressing digital preservation issues at all. For example, the simplicity and universality of printing to paper or microfilm come at the expense of great losses in the functionality of digital information. Migration strategies that involve reformatting of digital materials to a simple standard format usually eliminate the structure of documents and relationships imbedded in databases. Computation capabilities, graphic display, indexing, and other features often are lost, thus limiting future analytical potential. Normalization to standard formats is not always technically feasible and it usually is quite costly.

Archives and libraries must also contend with entirely new forms of electronically-enabled discourse and new forms of artistic and cultural expression that do not have predecessors in the analog world. No current preservation method is adequate for preserving dynamic data objects from complex systems. There are no established conceptual models or technical processes for preserving multi-media works, interactive hyper-media, on-line dialogues, or many of the new electronic forms being created today. The archival requirements to preserve content, context, and structure and to maintain the capability to display, link, and manipulate digital objects only heighten their software dependency.

The preservation community is only beginning to explore possible alternatives to storing digital information in "software- independent" form. Rothenberg (1995) proposed an approach for maintaining the content of digital materials intact without losing the ability to retrieve meaning-rich sources. He recommended retaining the original document in its original format encapsulated in a virtual "envelope" that contains software instructions for retrieval, display, and processing of the message in the envelope. The envelopes would contain contextual information and the transformational history of each object. Execution of the instructions would rely on an archive of hardware and software emulators or on instructions in the envelope with specifications to construct emulators.

Important research is underway to define standards for data interchange that can support electronic commerce and satisfy business requirements in a variety of environments. A major research project at the University of Pittsburgh is defining metadata requirements for evidence that will support the need for integrity, authenticity, reliability, and archiving through standards for "metadata encapsulated objects" (Bearman and Sochats). Archives, libraries, and other institutions with preservation responsibilities will benefit if systems are built to implement such metadata standards. Wide scale adoption of data and communication standards by the originators of digital information to support current business needs will also facilitate long-term preservation. Rapid implementation of electronic commerce depends on widespread development and adoption of standards for EDI (electronic data interchange) transaction sets under auspices of the ANSI X.12 committee, and many organizations are adopting standards for format and definitions to enable exchange, reuse, and sale of digital information and to reduce conversion and maintenance costs. Standards initiatives that address business needs for the secure and reliable exchange of digital information among the current generation of systems will impose standardization and normalization of data that ultimately will facilitate migrations to new generations of technology. Yet to benefit fully from the synergy between business needs and preservation requirements we have to move beyond simply paying lip service to cultural heritage concerns and recognize that equally critical social goals, such as long-term genetic research, monitoring global environmental change, locating nuclear waste sites, and establishing property rights are also dependent on long-term access to reliable, electronic evidence.

Much remains to be done in research, development, and implementation before we can assume that even a small portion of our digital heritage will survive more than a few years. It is fair to say that the state of development in digital preservation remains largely experimental. Only a few libraries, archives, and other institutions have established digital preservation programs, while most research and innovation comes from pilot projects and prototypes. Tested methods that have proven effective on a small scale in a limited number of repositories are not feasible for preservation of many of the types of digital materials that archives and libraries will confront in their preservation endeavors.

Areas for Research and Development

The current state of digital preservation suggests several fruitful areas for research and development. I will discuss four areas: storage media, migration, conversion, and management tools. These four domains are often mutually dependent and ultimately must to be integrated into an infrastructure for digital preservation. Yet better solutions are necessary in all four areas before such integration can occur. Finally, I will share some observations about the issues of scale and cost that must be considered if we are going to achieve any degree of systematic preservation.

Storage Media

The limited life of magnetic and optical media pose a significant problem, although this is not the primary limiting factor for digital preservation. Recent research on the longevity of magnetic media indicate a useful life span of 10 to 30 years if they are handled and stored properly. Some optical disk technologies promise life spans of up to 100 years. Most authorities argue that enhanced media longevity is of little value because current media outlast the software and devices needed to retrieve recorded information.

Nevertheless, improvements in the stability, capacity, and longevity of the base storage media are needed to drastically reduce the vulnerability of digital materials to loss and alteration and to lower storage costs. Ample research and experience provide evidence of what can go wrong with magnetic media as a result of binder degradation, magnetic particle instabilities, and substrate deformation (Van Bogart). Optical media are susceptible to damage from high humidity, rapid and extreme temperature fluctuations, and contamination from airborne particulate matter (U.S. National Archives and Records Administration). To prevent these problems, it is imperative to store magnetic and optical media under strict environmental controls that are not always available, affordable, or convenient. Even modest improvements which produce storage media with larger per unit storage capacities and greater tolerance to variations in temperature and humidity will lower preservation costs by lessening the need for strict environmental controls, reducing the frequency with which digital media must be "refreshed" through recopying, and decreasing the number of storage units that must be handled.

This raises the question, however, of whether research on incremental improvements in current storage technologies will benefit preservation in the long run or whether we should seek alternative approaches to digital storage that more adequately meet archival requirements. As a frame of reference it is worth remembering that microfilm, which is considered the only acceptable archival storage medium, lasts at least 300 years with minimal maintenance if stored properly. Last June, the Los Alamos National Laboratory announced the invention of a High-Density Read-Only Memory (HD-ROM) technology that uses an ion beam to inscribe information on pins of stainless steel, iridium, or other materials. The HD-ROM is capable of storing 180 times more information than current CD-ROM technology at roughly one-half percent of CD-ROM costs. According to the release about this technology, the HD-ROM is impervious to material degradation and it requires no bit stream interpreter because the technology can describe in human-readable form all of the instructions needed to interpret the data (LANL Ion Beam Storage). Such an approach illustrates the potential for solutions built on entirely new storage technologies.

Migration

Better methods for migration of digital materials to new generations of hardware and software are much needed for digital preservation regardless of breakthroughs in mass storage technologies. Planning for migration is difficult because there is limited experience with the types of migrations needed to maintain access to complex digital objects over extended periods of time. When a custodian assumes responsibility for preserving a digital object it may be difficult to predict when migration will be necessary, how much reformatting will be needed, and how much migration will cost. There are no reliable or comprehensive data on costs associated with migrations, either for specific technologies and formats or for particular collections, and little research underway on methodologies that would reduce the costs and burdens of migration.

The preservation community as a whole would benefit tremendously from the development of backward compatibility paths that would be included as a standard feature of all software. Backward compatibility or migration paths would enable a new generation of software to "read" data from older systems without substantial reformatting and without loss of retrieval, display and computational capabilities. Although backward compatibility is increasingly common within software product lines, migration paths are not commonly provided between competing software products or for products that fail in the marketplace.

Stewards of digital material have a range of options for preserving digital information. One might preserve an exact replica of a digital record with complete display, retrieval, and computational functionality, or a representation of the record with only partial computation capabilities, or a surrogate for the record such as an abstract, summary, or aggregation. Detail or background noise might be dropped out intentionally through successive generations of migration, and custodians might change the format or storage media. Enhancements are technologically possible through clean-up, mark-up, and linkage, or by adding indexing and other features. These technological possibilities in turn impose serious new responsibilities to present digital materials to users in a way that allows them to determine the authenticity of the information and its relationship to the original record. Methods to document changes in digital objects during their life span need to be incorporated as an integral part of improved migration methods.

There are few well developed methods for preserving and migrating software so that it might be used to recreate digital documents that have the "look and feel" of the original sources. Maintaining repositories of obsolete hardware and software has been discussed periodically, but usually dismissed out of hand as too expensive and not demonstrably feasible. This approach deserves more serious consideration as a strategy for maintaining continuing access to certain types of digital materials. Feasibility studies and cost/benefit analyses should be conducted to determine the technological, economic, and commercial feasibility of maintaining selected legacy software systems and performing specialized migrations or, alternatively, of building and maintaining software emulators. Such an approach would support replay of original sources and contribute to the preservation of software as a significant cultural and intellectual resource in its own right.

Conversion

Faster, cheaper, and higher resolution conversion technologies are another critical element needed to make digital preservation feasible on a large scale. Most archivists and librarians accept the fact that we live in a hybrid environment where paper, microfilm, video, and magnetic and optical media need to interoperate in a more integrated and transparent manner. The vast majority of primary sources today still reside on paper and/or microfilm with little chance that we will see the mass conversion of existing archival and library holdings to digital formats. Research and planning for digital preservation must recognize that we will be dealing with conversion for a long time and that investments in improving capture rates, accuracy, resolution, and verification will have long-term benefits. Moreover, improvements in conversion technologies may support hybrid solutions to preservation and access problems by permitting repositories to store certain formats of digital material on stable media, such as microfilm, with on demand conversion to digital form for analysis and reuse. Efforts to capture and store descriptive mark-up on film for subsequent conversion are hampered by unacceptable error rates in OCR technology and cumbersome conversion processes (Giguere).

Management Tools

A fourth area for research is in the development of management tools for digital libraries and archives that integrate descriptive control and maintenance with storage technologies. Dynamic digital objects, such as those found in hypertext systems, pose special management problems for both current and future retrieval and reuse. The boundaries of hypertext sources, like those found on the World Wide Web today, are difficult to ascertain because no single party or institution controls changes in the nodes and links that make hypertext objects live and highly responsive information resources. A high degree of volatility accompanies these objects because the contents of nodes change, the sites where information resources are stored change, and the links between nodes change, move, and vanish. Some recent tools, such as the MOMspider (Multi- Owner Maintenance Spider) and Web:Lookout are capable of traversing a portion of the Web and noting maintenance problems such as broken links, moved documents, modified documents, and objects that have exceeded their expiration dates (Ackerman and Fielding). While tools such as these are useful for current maintenance, they do not address long-term preservation concerns. If further developed to address preservation problems, tools such as these have the potential to serve as filters, identify similar or identical objects, and monitor for maintenance problems.

Research and development of tools that would imbed more intelligence about the preservation status of digital material into the objects themselves would make monitoring and maintenance of large digital collections more automatic. Current methods for monitoring the physical status of digital materials are labor intensive, unreliable, and potentially damaging to the materials themselves. Recommended procedures for monitoring physical deterioration of magnetic tape, for example, involve reading a small sample of tapes periodically to determine whether any data losses have occurred (Eaton). The potential exists to build monitoring and reporting mechanisms into digital objects, storage systems, and network architectures that could support self- reporting of physical status and initiate automatic maintenance procedures.

Despite differences, some lessons from traditional preservation are transferable to the digital environment. In order to avoid commitments that far exceed available resources and costly rescue and restoration efforts, preservation must become an integral part of the planning, design, and resource allocation for digital libraries and archives. Integration of preservation requirements and methods with access and maintenance systems is essential to fully and efficiently support the processes of migration, regeneration, and documentation of the life of digital objects. Planning for preservation must become an integral part of the design and management of digital libraries and archives. If left as an afterthought, there is little reason to believe that long-term preservation of digital information will be any more affordable than preservation of conventional formats has been.

Some concluding comments

In closing, I would like to mention three issues that I discussed only in passing. One concern is the question of scale and scalability. The preservation community has at its disposal a variety of tactics for digital preservation that appear to work effectively for certain types of materials in certain restricted environments, but we have not yet developed solutions that are scalable to the general problem of digital preservation. This is not to suggest that there is or should be a single solution to digital preservation. The methods used will vary depending upon the complexity of the original data objects, the extent to which the functionality for computation, display, indexing, and authentication must be maintained, and the requirements of current or anticipated users. But any solution must be scalable from the laboratory, prototype, or pilot project to the wide range of individuals and institutions who have a need to make digital materials last longer than the current generation of technology permits.

Another closely related issue is the question of affordability. Regardless of how the responsibility for digital preservation is distributed, societies only allocate a small and finite amount of resources to preserving scholarly and cultural resources. And in the digital environment it seems likely that more preservation responsibilities will be distributed to individual creators, rights holders, distributors, small institutions, and other players in the production and dissemination process. Therefore, it seems imperative that digital preservation technologies become affordable and accessible to the wide range of individuals and institutions that will attempt to preserve digital materials.

Finally, it would be beneficial to both the preservation community and to those conducting research on issues of longevity, migration, and conversion if there were more venues for exchange of ideas, requirements, and recent developments. Without a continuing dialogue between humanists, preservationists, and the scientific community it is difficult to include preservation requirements in scientific research endeavors, and it is challenging for those of us outside the scientific community to keep up with and evaluate new products. I hope that the discussion we are beginning at this conference will lead to more regular and formal processes for linking the needs of scholars and preservationists with the research agendas and projects of scientists.

References

Ackerman, Mark S. and Roy T. Fielding. (1995). "Collection Maintenance in the Digital Library," URL: .

Bearman, David and Ken Sochats, (1995). "Metadata Requirements for Evidence," Draft, maintained on the WWW server for the University of Pittsburgh Project, Functional Requirements for Recordkeeping, URL: .

Conway, Paul. (1994). "Digitizing Preservation." Library Journal, (February 1, 1994): 42-45.

Conway, Paul. (1990). "Archival Preservation in a Nationwide Context," American Archivist, 53, No. 2: 204-22.

Eaton, Fynnette L. (1993). "The National Archives and Electronic Records For Preservation," in Preservation of Electronic Formats: Electronic Formats for Preservation, Janice Mohlhenrich, ed., Ft. Atkinson, WI: Highsmith Press: 41-61.

Giguere, Mark D. (in progress). "Electronic Document Description Standards: A Technical Feasibility of Their Use in the Augmentation of the Microform Preservation of Contextual Cues Embedded in Structured Electronic Documents During Successive Digital/Analog/Digital Reformatting," dissertation proposal submitted to the School of Information Science and Policy Studies, State University of New York at Albany, (January 5, 1995).

Gould, Constance. (1988). Information Needs in the Humanities: An Assessment, Stanford, CA: The Research Libraries Group.

Graham, Peter S. (1995). "Requirements for the Digital Research Library," College and University Research Libraries, July 56, No. 4: 331-39.

"LANL Ion Beam Storage Holds 180 Times More Info than CD-ROMS," Science and Engineering News, June 23, 1995, down loaded from HPCwire and redistributed to .

Lesk, Michael. (1992). Preservation of New Technology: A Report of the Technology Assessment Advisory Committee to the Commission on Preservation and Access. Washington, D.C., Commission on Preservation and Access.

Levy, David M. and Catherine C. Marshall. (1995). "Going Digital: A Look at Assumptions Underlying Digital Libraries," Communications of the ACM, 58, No. 4: 77-84.

Lynch, Clifford. (1994). "The Integrity of Digital Information: Mechanics and Definitional Issues," Journal of the American Society for Information Science, 45, No. 10: 737-44.

Michelson, Avra and Jeff Rothenberg. (1992). "Scholarly Communications and Information Technology: Exploring the Impact of Changes in the Research Process on Archives," American Archivist 55, No. 2: 236-315.

O'Toole, James M. (1989). "On the Idea of Permanence," American Archivist, 52, No. 1: 10-25.

The Preservation of Archival Materials. (1993). Washington, D.C.: Commission on Preservation and Access.

Preserving The Intellectual Heritage: A Report of The Bellagio Conference. (1993). Washington, D.C.: The Commission on Preservation and Access.

Rothenberg, Jeff. (1995). "Ensuring the Longevity of Digital Documents," Scientific American, 272 No. 1: 24-29.

Task Force on Archiving of Digital Information. (1995). "Preserving Digital Information," Report of the Task Force, commissioned by the Commission on Preservation and Access and The Research Libraries Group, Version 1.0, August 24, 1995.

U.S. National Archives and Records Administration. (1994). Digital Imaging and Optical Digital Disk Storage Systems: Long-Term Access Strategies for Federal Agencies. Technical Information Paper No. 12. National Technical Information Service, Washington, D.C. URLs: and .

University of the State of New York, State Education Department, State Archives and Records Administration. (1995). Building Partnerships for Electronic Recordkeeping: The Final Report and Working Papers of the Building Partnerships Project. Albany, NY.

Van Bogart, John W. C. (1995). Magnetic Tape Storage and Handling: A Guide for Libraries and Archives, Washington, D.C.: Commission on Preservation and Access and the National Media Laboratory.