Published on:

Digital Access to Archival Works: Could 108(b) Be the Solution? by Peter Hirtle

Digital Access to Archival Works: Could 108(b) Be the Solution?

By Peter B. Hirtle [*]

September 24, 2006


Section 108(b) of the Copyright Law, which deals with unpublished works, is often described primarily has a “preservation” clause, with its primary purpose being to ensure that our manuscript heritage is not lost. A closer look at the legislative history of the section, however, reveals that Congress was primarily concerned with increasing scholarly access to unpublished materials. Limited distribution to other libraries and archives to enhance research access to the original materials, it concluded, does not compete with the copyright owner’s right to commercially exploit the work. Under the original section 108(b), there were no limits on the number of copies that could be made for deposit in other repositories. Today digital technologies could provide a means of providing access to research materials without having to distribute physical copies to other repositories (though distribution of copies for preservation purposes would still be desirable).


Section 108 of the Copyright Law specifies some of the specific copyright exemptions available to libraries and archives that limit the scope of the copyright owner’s exclusive rights. One section, 108(b), focuses specifically on unpublished works. It specifies that it is not an infringement of copyright when an eligible library or archives makes “three copies or phonorecords of an unpublished work duplicated solely for purposes of preservation and security or for deposit for research use in another library or archives.”

Section 108(b) is most often viewed by analysts as a “preservation” exemption. Lolly Gasaway, for example, describes Section 108(b) as “a true preservation section,” in contrast to 108(c), which only authorizes the replacement of some published items.[1] The American Library Association in its analysis of the Digital Millennium Copyright Act concluded that the library distribution rights found in Section 108 were limited to interlibrary loan. The purpose of the reproductions made under Section 108(b) and (c), therefore, was for preservation.[2]

When one examines the original legislative rationale for Section 108(b), however, it becomes rapidly apparent that access, and not preservation, was the driving impetus for inclusion of the section in the 1976 Copyright Act. Congress wanted to ensure that scholars had ready access to unpublished research materials. Such access could be secured by authorizing the distribution of copies of the unpublished research materials to other libraries and archives. Congress apparently accepted the arguments that limited distribution of facsimiles of unpublished manuscripts would in no way conflict with the publication and distribution rights of the copyright owner; scholarship and commercial distribution are not in conflict.

With the advent of ubiquitous network access to digital information, and in the light of the original intent of Section 108(b), it is easy to conclude that limited networked digital access to unpublished material would better address the problems that Congress tried to solve in 1976.

The Origins and Intent of Section 108(b)

Section 108(b) of Title 17 permits the reproduction of unpublished material for “preservation and security or for deposit for research use in another library or archives.” In order to think about this provision in the digital era, it would be helpful to understand what problem Congress thought it was solving when it included this provision in the 1976 Act.

Mary Rasenberger and Chris Weston’s excellent background paper on Section 108 has this to say about the provision:

In the same report, the Judiciary Committee also added a new provision, urged in 1965 by the General Services Administration, historians, archivists, and educators, on reproduction of works in archival collections. It was the first iteration of the current section 108, and provided:

Notwithstanding the provisions of section 106, it is not an infringement of copyright for a nonprofit institution, having archival custody over collections of manuscripts, documents, or other unpublished works of value to scholarly research, to reproduce, without any purpose of direct or indirect commercial advantage, any such work in its collections in facsimile copies or phonorecords for purposes of preservation and security, or for deposit for research use in any other such institution.

The explanation of this provision, to which the committee noted there was “little or no opposition,” said it would not permit archives to make machine-readable copies, to distribute the copies to scholars or the public, or to override prior contractual arrangements.[3]

While providing very useful background information, the quote from the House Report does not really offer an explanation of the provision. For that, we need to look at why the General Services Administration introduced the provision and the rationales that were provided in support of it.

The initial suggestion for the provision was made by Dr. Robert H. Bahmer, Deputy Archivist of the United States, who testified before Subcommittee No. 3 of the House Judiciary Committee on 17 June 1965.[4] Bahmer noted that the proposed new general copyright law would have two major detrimental effects. First, it would hurt the efforts of the National Historical Publications Commission (NHPC), a recently authorized component unit of the National Archives and Records Service (NARS). One of the major functions of NHPC was to give grants to State and local agencies to microfilm collections of nationally significant manuscripts in order that these materials could be available for historical research.[5] Bahmer worried that the proposed copyright legislation would be a real deterrent to this work.

More importantly, Bahmer cited the general harm to historical scholarship that the draft copyright legislation would cause. “The bill,” Bahmer noted:

would virtually preclude the microcopying of collections of papers by a historical society or other custodian for the purpose of depositing the copy in another institution unless the documents involved are at least 100 years old. It is not even clear whether…a custodian could microfilm or otherwise copy a collection of papers less than 100 years old when the purpose is to provide an insurance copy in case the originals are lost as a result of some natural or man-made catastrophe.[6]

Bahmer added that the bill would also prohibit the reproduction of historically important documents found in one section of the country “in order to make copies of them available for study in another section.”[7]

As a consequence, Bahmer proposed that the bill be amended to permit reproduction of manuscript collections for historical research purposes. The provision, Bahmer asserted, should make it clear that it is not an infringement of copyright to reproduce all or parts of a manuscript collection “for purposes of preservation and security or for the deposit of copies thereof in other such institutions for research use, provided that the copies are made available without any motive of financial gain.”[8]

GSA had prepared, and presented to the committee, draft language for a new section in the copyright law to accomplish this goal. The proposed language is worth quoting in full:

(a) Notwithstanding the provisions of section 106, it is not an infringement of copyright for a library, historical society, archives, manuscript repository or other educational or research institution having legal custody of manuscript collections of business, personal, and institutional papers and other documents which are of value to historical research, to reproduce such collections, or parts thereof, in microfacsimile form:

(1) for purposes of preservation and security, or

(2) for the deposit of copies in other institutions in order to make the manuscript collections more widely available for research use, if the copies made are without any purpose of financial gain.”

(b) The exclusive rights of the owners of the copyright in the manuscript collections shall extend to all copies thereof made under the provisions of this section.

(c) As used in this section the term ‘manuscript collections’ shall include both ‘literary works’ and ‘pictorial, graphic and sculptural works’ as defined in section 101 of this act.[9]

The House committee’s questions to Bahmer centered on two primary issues. One was the meaning of “microfacsimile.” Bahmer and the other GSA representatives indicated that it would include any reproduction in reduced format, but would exclude such things as Thermofax reproduction, which are the same size as the original. No explanation is given as to why they wanted to restrict the reproductions to reduced sizes. [10]

The second set of questions centered on the use of unpublished material. In his responses to Representative Richard C. Poff, Bahmer made clear that he considered that there were two types of uses of manuscript materials. The first type of use, publication, required, and would continue to require, the permission of the copyright owner. The second type of use, making copies for research purposes, however, should be exempted. The unpublished material could be projected onto a screen or a reader, but Bahmer indicated that it would not be permissible to make a print from the microfacsimile. In this way, no substantive rights of the copyright owner would be affected.[11]

Charles W. Gasque, the Deputy General Counsel for GSA, added that under the proposed provision, the use the researcher could make of the material would be the same as in the repository where the original was located. If you could review, read, or look at a manuscript at the University of Virginia, you should be able to do the same with a copy of the manuscript given to the University of Chicago. The provision, Gasque noted, was needed as a matter of convenience to make it easier for historians to conduct historical research without having to do extensive travel. Convenience would not be served if only one copy were deposited somewhere else.[12]

Support for the idea proposed by Bahmer and GSA was strong. The Council of the Society of American Archivists (SAA) had already adopted a resolution on the new copyright bill that may have been the stimulus for Bahmer’s testimony.[13] It recommended in part:

That the new copyright law include a provision that will permit libraries, archives and other repositories to microfilm or otherwise make facsimiles of manuscripts for the purpose of depositing copies in other institutions.

This would not adversely affect copyright interests in the manuscripts in any way. It would, in effect, merely extend the repository’s ability to show its manuscripts to scholars and others, and, at the same time, would make it possible to house security copies at a distance from the originals.[14]

The resolution, as endorsed by the National Trust for Historic Preservation, the Organization of American Historians, the Western History Association, and the Harry S. Truman Library Institute, was presented to Congress during the testimony of Professor Julian Boyd. Boyd, who testified on behalf of many of these organizations as well as the American Historical Association and the American Association for State and Local History, spoke out strongly in favor of the proposed new provision. Scholars, Boyd noted, need access to scholarly resources wherever they may be. He added one caveat: Boyd objected to limiting the proposed exemption to microfacsimiles. He believed that the community of scholars would insist that the exemption be extended to copies in any format. “To compel the scholar to depend exclusively on microfacsimile reproductions in the narrow sense given to the term in the preceding testimony,” Boyd concluded, “would be to curtail the immense possibilities now before us of increasing access to these resources throughout the country and wherever scholars may happen to be.”[15]

Only two other groups spoke out on the proposed new provision (though the subcommittee did publish the three letters it had received on the new provision, all of which supported the SAA resolution). The first was the American Council on Education. In his testimony, its representative noted that the Council supported the recommendation of the Deputy Archivist of the United States that provision be made for the duplication of unpublished manuscripts “for library and scholarly research purposes.”[16]

More importantly, the Copyright Office also spoke up on behalf of the GSA provision. Testifying on behalf of the Office, Abraham L. Kaminstein, the Register of Copyrights, noted that he:

was strongly impressed by the arguments advanced by the Deputy Archivist of the United States and by Prof. Julian Boyd on behalf of archivists and historians. Their proposal that libraries, archives, and other repositories be permitted to duplicate manuscripts for the preservation of their own collections and for research use in other archival institutions struck me as reasonable and worthy of adoption.[17]

In its 1966 report on the revision bill, the House Judiciary Committee accepted in spirit the provision proposed by GSA and endorsed by the Register of Copyrights. While noting that it was not sympathetic to the introduction of a general exemption for library and archives as some had proposed, it had been convinced by the arguments that it had heard about the difficulty of conducting archival research. The committee noted that arguments had been made expressing the need to be able to make facsimile reproductions of unpublished works for deposit in other manuscript collections. Proponents noted as well that unpublished material in archives is of little interest to copyright owners, but of great historical and scholarly value. They also noted that a limited right to duplicate archival collections “would not harm the copyright owners’ interests but would aid scholarship and enable the storage of security copies at a distance from the originals.”[18]

The Committee described its response to these recommendations as “generally sympathetic.” It therefore proposed that the law include a new section to authorize these activities: Section 108.[19] An institution could take advantage of the new section if the copies made were either for “preservation and security” or for “deposit for research use in any other such institution;” this provision was not intended to authorize copies for users. The Committee rejected the GSA’s initial suggestion to limit the copies to micofacsimiles, but did accept some of the concerns raised at the hearings in other contexts about the potential risks associated with storing copyrighted works in electronic form by distinguishing between facsimile reproduction, which would be permitted, and reproduction “in ‘machine-readable’ language for storage in an information system,” which would not be allowed.

Four things are clear from a review of the introduction and discussion surrounding the first appearance of what would become Section 108(b). First, the law as finally passed by Congress in 1976 mirrors almost exactly the language proposed by GSA in 1965. In theory, Congress could have decided to use this language for reasons that had nothing to do with rationale presented to them by the provision’s supporters, but it is much more likely that in accepting the language, Congress also accepted the provided rationale for the provision.

Second, the driving impetus for the suggested provision was increased scholarly access, not preservation. Scholarly access could be enhanced through the distribution of copies of manuscript collections to other institutions. Preservation was important, but it remained secondary in the discussions (in spite of its primary position in the proposed legislation). The primary goal of Section 108(b) was to make unpublished research material broadly available to the scholarly community.

Third, there is no suggestion in the proposed legislation or in the reported discussion that the number of copies of manuscripts deposited in other libraries should be limited. As Gasque noted during GSA’s testimony, the provision was intended primarily as one of convenience to make it easier for historians to conduct historical research without having to do extensive travel. Convenience would not be served if only one copy were deposited somewhere else. The Committee’s own report talks about “facsimile copies” in the plural, whereas if they intended only one copy to be made, they could have easily have used the singular.

Fourth, while the provision was initially proposed by GSA and NARS, the real need for the provision was found in other types of institutions the “library, historical society, archives, manuscript repository or other educational or research institution” suggested in GSA’s draft provision. The National Archives had then, and still has, statutory exemption from copyright infringement with regard to unpublished, unregistered items.[20] Section 108(b) should not be viewed, therefore, as a provision to aid the National Archives. It was, instead, intended to assist all other historical repositories by extending to them just one of the broad privileges available to the National Archives.

From Proposal to Enactment

Section 108 as proposed by the House Judiciary Committee remained in the House and Senate versions of draft copyright legislation through the 1960s. It sparked little controversy or debate in subsequent hearings. One of the few times that the proposal was mentioned specifically was in 1967 when Arthur R. Miller, then of the University of Michigan Law School and speaking on behalf of EDUCOM, faulted the insistence in the proposal on facsimile reproductions. The exclusion of machine reproduction, he argued,

represents pure fear of new techniques that is not based on any showing of actual injury to the owner of the archived material. If an archived manuscript can be reproduced on a Xerox machine 100 times and distributed to 100 other nonprofit libraries under section 108, why cannot exactly the same thing be done by allowing this same manuscript to be put into machine-readable form, input into a data processing system, and sent to 100 terminals in the same 100 nonprofit libraries?[21]

The text remained the same until 1969 when the Senate Committee Print slightly modified the language to that which was adopted in 1976.

Section 108(b) and the DMCA

In 1998, with the passage of the Digital Millennium Copyright Act, Section 108(b) was slightly amended. Two important changes were made. First, the number of copies that could be made was limited to three. This was in all likelihood an unintentional change. Prior to the passage of the DMCA, Section 108(c), which governs the making of replacement copies of published works, had specified that a single facsimile copy could be made. Preservation microfilming standards, however, require that when filming a work, at least three copies of the work should be made: a camera negative; a print master; and a service copy. Library groups were successful in their efforts to suggest that the language for 108(c) should be changed to reflect current practice. No representatives of the archival community, however, were present at the “hastily called” session with Senate committee staff at which the changes were worked out.[22] As a result, the same language that was proposed to increase the scope of Section 108(c) was applied to Section 108(b), restricting for the first time the limited distribution right passed by Congress in 1976.

Second, the restriction that the copies made under Section 108(b) and (c) be in facsimile form was dropped and digital copies were permitted. While it could be argued that “facsimile” was a technologically neutral term and could in theory apply to digital scans of unpublished items, in practice most commentators (including the Copyright Office) assumed that “facsimile” meant “analog.”[23] As we have seen, when legislators were crafting Section 108, they worried that the little-understood practice of storing texts in machine-readable form could be an undue imposition on the rights of the copyright owner. By 1998 it had become clear that a scan of a page in book was little different than a photocopy. Because some perceived that the danger of systematic reproduction and distribution was higher with digital copies, however, use of these digital copies was limited to the physical premises of the library. Again, the rules that were developed for replacement copies were applied to unpublished materials as well.

Section 108(b) in the Digital Era

In an era of ubiquitous digital technology, how might Section 108(b) be altered to better address original Congressional intent?

First, as we have seen, the primary purpose of Section 108(b) was to provide convenient access to the vast array of unpublished resources of possible scholarly interest found in the historical repositories of the nation. At the time that Section 108(b) was drafted, the only practical way of providing access to material at a distance was via depositing microfilm or photographic copies at another institution. There are, however, limitations with the technology. For example, some important research materials do not lend themselves to effective microfilming. They may have material that is in color, or a large number of faint marginalia.

There are also scholarly risks associated with using a copy of a research collection made years or sometimes decades earlier. For example, a collection that has been microfilmed can still change and evolve, even after filming. New documents can be added to the collection, and other documents that were filmed may at a later date be identified as forgeries. The original, authoritative collection as found in the owning repository can be updated and corrected, but a microfilmed collection must by its very nature remain static.

In addition, the repository that has ownership of the originals may also place requirements on their use. These requirements are likely to change over time. For example, the policies of the institution may change, or restrictions based on privacy or other factors may disappear over time. The repository that owns the original items will change access to reflect current policies, but the repository that owns the microfilm has less incentive to track changes in access and use policies at the primary repository. The best access controls are the ones managed by the repository that owns the originals.

Digital technologies can address the goals of the original legislation and overcome the limitations of microfilm. Section 108(b) should be altered to allow archival and manuscript repositories to reproduce and deliver historical research material in digital form. The advantages of this approach are obvious:

  • The increased scholarly access to research materials that is at the heart of Section 108(b) is enhanced.
  • The repository that owns the material, and hence is in the best position to convey information about any access and use restrictions, becomes responsible for administering access.
  • The issue of how many copies is it reasonable to distribute to other archival repositories becomes moot.

In order to respect the copyright owner’s right of first publication and any commercial interests that there may be in the unpublished material, two important provisions should be added and a third considered:

(1) While there may be transcriptions or other machine-readable versions of the documents available for indexing purposes, the documents presented to the user must be images of the original pages. (Note: Behind this recommendation is the belief that no online compilation of facsimile documents would likely damage a market for an edited, annotated version of the original documents.)

(2) Nothing in this section in any way affects any contractual obligations assumed by the library or archives when it obtained a copy or phonorecord of a work in its collection, nor does it affect any other legal interests such as privacy and publicity rights.

(3) The vast majority of repositories have procedures in place for registering readers before they can work with research materials. It may be desirable to require similar registration before users are allowed access to online unpublished material. Such a registration system would be a minor imposition on the user. It would, however, calm the fears of those who might worry that the digital availability for scholarly purposes of the unpublished material is the equivalent of general publication. It would also provide the opportunity for the archives to inform all users that access if for scholarly or research purposes only, and that other uses may require the permission of the copyright owner.

In addition to allowing repositories to post page images of unpublished material for access purposes, Section 108(b) will still need to allow repositories to distribute material to other institutions as part of a distributed preservation system.


When what was to become Section 108(b) of the Copyright Act was first proposed, the storage of unpublished materials in digital form was perceived to be a threat. With thirty years of experience with distributed archival collections, digital technologies, and networked access, it is now apparent that digital technologies are the best way of addressing the challenge of access to unpublished materials. The distribution of physical copies of unpublished materials to other repositories is still needed to ensure the preservation and security of the information. However, the best method of providing scholarly access to unpublished materials, which was the primary motivation for the inclusion of Section 108(b) in the Copyright Act, is via networked digital access from the owning institution.

* © 2006 by Peter B. Hirtle, Intellectual Property Officer, Cornell University Library, This work is made available under a Creative Commons Attribution-NonCommercial-ShareAlike 2.5 license. To view a copy of this license, visit or send a letter to Creative Commons, 543 Howard Street, 5th Floor, San Francisco, California, 94105, USA. In addition, permission is granted for non-profit educational use, including but not limited to reserves and coursepacks made by for-profit copyshops.

[1] Laura N. Gasaway, “America’s Cultural Record: A Thing of the Past?,” 40 Hous. L. Rev. 643 (2003). Available by clicking here.

[2] Carol C. Henderson, American Library Association, Washington Office, “Library Preservation: Changes Incorporated in H.R. 2281 The Digital Millennium Copyright Act of 1998 (PL 105-304),” at

[3] Mary Rasenberger and Chris Weston, “Overview of the Libraries and Archives Exception in the Copyright Act: Background, History, and Meaning,” p. 15. Available at

[4] Technically, Bahmer was testifying on behalf of the Administrator of the General Services Administration. The National Archives and Records Service was administratively a part of GSA; it would not regain its status as an independent agency until 1984, when its named was also changed to the National Archives and Records Administration.

[5] While he did not mention it, microfilming in order to increase research accessibility might also have the secondary benefit of preserving the unpublished materials.

[6] United States. Congress. House. Committee on the Judiciary. Copyright law revision. Hearings before Subcommittee no. 3 of the Committee on the Judiciary, House of Representatives, Eighty-ninth Congress, first session. (Washington: Government Printing Office, 1966): 1110.

[7] Ibid., 1111.

[8] Ibid., 1112.

[9] Ibid., 1114.

[10] Two possibilities for the limitation to “microfacsimiles” come to mind. The first is that it may never had occurred to the NARS officials that a repository would want to make a full-size reproduction of a collection for deposit in another archives. Microfilm was the technology currently in use at NARS and by the NHPC. The lack of any opposition to later efforts to remove the limitation might suggest this is the case. Alternatively, NARS officials might have viewed microfasimiles much as thumbnails are viewed today: as low resolution surrogates that could never replace the originals. Further investigation into the records of the GSA and NARS (now the National Archives and Records Administration) would be needed in order to determine if either argument is correct.

[11] Ibid., 1116. Bahmer’s analysis of what should be legally acceptable may have been shaped by the available technology. At the time he testified, there was no easy way to produce hard copy prints from microfilm copies. Scholars would, one assumes, still be able to make complete transcriptions of the documents on the screen, just as they could from the originals.

[12] Ibid.

[13] Bahmer was also chair of SAA’s Ad Hoc Committee on Copyright that drafted the resolution.

[14] Society of American Archivists, “Minutes of Council, 23 April 1965.” American Archivist 28:3 (July, 1965): 467-8.

[15] Ibid., 1143.

[16] Ibid., 1564.

[17] Ibid., 1867.

[18] House hearing, p. 66.

[19] The text of the section as proposed by the Committee is quoted in full by Rasenberger and Weston, supra at supra n. 3.

[20] 44 U.S.C. § 2117. Limitation on liability, currently reads as follows:

When letters and other intellectual productions (exclusive of patented material, published works under copyright protection, and unpublished works for which copyright registration has been made) come into the custody or possession of the Archivist, the United States or its agents are not liable for infringement of copyright or analogous rights arising out of use of the materials for display, inspection, research, reproduction, or other purposes.

[21] United States. Congress. Senate. Committee on the Judiciary. Copyright law revision. Hearings before Subcommittee on Patents, Trademarks, and Copyrights of the Committee on the Judiciary, United States Senate, Ninetieth Congress, first session, pursuant to S. Res. 39 on S. 597. (Washington: Government Printing Office, 1967): 560.

[22] Carol C. Henderson, Library Preservation: Changes Incorporated in H.R. 2281, The Digital Millennium Copyright Act of 1998 (PL 105-304) (Washington, DC: American Library Association Washington Office, Nov., 1998). Available at

[23] See, for example, the testimony by Arthur Miller cited in n. 23; Library Reproduction of Copyrighted Works (17 U.S.C. 108), Report of the Register of Copyrights (January 1983); and Robert L. Oakley, Copyright and Preservation: A Serious Problem in Need of a Thoughtful Solution (Washington: Commission on Preservation and Access, Sept. 1990). Available by clicking here.