March 16, 2007
Policy Advisor for Special Programs
U.S. Copyright Office
Dear Ms. Rasenberger:
Thank you very much for the opportunity to comment, both in person and in writing, on the important issues being considered by the Section 108 Study Group.
We are submitting the following written comments to the Section 108 Study Group on behalf of the American Association of Law Libraries (AALL).
Jonathan A. Franklin
Chair, Copyright Committee
American Association of Law Libraries
Mary Alice Baish
Associate Washington Affairs Representative
American Association of Law Libraries
AALL Response to the Section 108 Study Group Regarding Interlibrary Loan and Other Copies for Users
March 16, 2007
The American Association of Law Libraries appreciates this opportunity to respond to the questions posed by the Section 108 Study Group on copyright exceptions for libraries and archives. The section headers below are keyed to the questions asked for the January 31st meeting.
Section 108 was written at a time when the typical interlibrary loan (ILL) transaction involved borrowing a monograph or a journal article in print form. Although print monographs and articles continue to be popular objects of ILL, it is much easier for most libraries to deliver a digital copy of an article to the user. The digital copy may or may not have begun as a print copy scanned into digital form, but after delivery to the user, no copy is retained by the borrowing Library. The Library does not know what happens to the downstream copy that goes to the user, but electronic ILL management systems usually provide the user with a warning that the material being provided has copyright restrictions.
It should be noted that the article may be borrowed for a variety of reasons, e.g., the Library does not own the journal issue; or the Library does own it, but the issue is either missing or has been sent to the bindery. The borrowing Library often has rights to the material, but the copy it owns is unavailable. Law libraries are different from some other types of libraries, in that we often own and/or access the same content in multiple formats and from multiple vendors, so there is often a licensed source that explicitly permits digital copying for in-house users as well as for ILL. Law libraries routinely add such a provision to licensing agreements if it is not already included.
Much of the content needed by law libraries is already in the public domain due to its age or the fact that it is a federal government document, and copyright protection is not at issue. Finally, when recent monographs are requested through ILL, it is much easier and less expensive for both the borrowing library and the lending library if the volume itself is sent, rather than a digital copy.
Electronic ILL management systems help libraries track the number of times they either borrow or lend a particular item, and ILL borrowing lists are routinely used as acquisition tools. Libraries would be unable to properly serve their users if required to provide only a print copy and ignore the benefits of better technology. Library users expect service whether or not they are able to visit the Library building, and better technology has made libraries much more able to serve the user who is handicapped, and the working student, as well as the user who has few options for transportation.
A.1. Currently, law libraries take full advantage of the Copyright Act (the Act), particularly Sections 107, 108, and 109, to provide lawful access to information resources. The Act has built-in safeguards to balance the interests of authors, users, and owners of copyrighted information. The most helpful provisions are format/medium neutral.
Even if something is "reasonably available at a fair price," sections 107 through 121 of the Act clearly support limitations to the rights granted in Section 106. Within fair use analysis, the fourth factor is still only one factor of the four enumerated factors. The American Association of Law Libraries feels strongly that the 1976 Act embodies the balance mandated by the Constitution, and that tinkering with the Act without considering the intended balance as Section 108 is currently written would be both unwise and distorting.
The balance struck in Section 108 permits copying of articles when users request those articles. This right cannot be revoked merely by making the articles available online for a fee. The underlying goal of 108 is to facilitate the spread of user-requested information while acknowledging the copyright holders' Section 106 rights.
Section 108 as currently written meets three major goals: equalization of information access, embodiment of Constitutional intent regarding information access, and a reiteration that the receiving library is merely standing in the shoes of the user.
It is clear that interlibrary loan is an equalizer, in that poor users with under-funded libraries are not disadvantaged when it comes to access to information. In a world without interlibrary loan, poorer users would have no way to gain access to the information they need. Due to the explosion in domestic and foreign legal publishing, not even Harvard Law Library can afford to acquire all works in the area. In a digital environment, this becomes even truer, as there may be no physical copy to lend. The balance struck in Section 108 equalizes access without eliminating the value of the Section 106 rights.
Section 108 embodies the balance between the incentive to create and the need for the public to have access to information. To reduce or eliminate the Section 108 right would violate the very goals of the Constitutional roots of copyright law. Section 108 permits libraries to stand in the shoes of their users when requesting information, thereby permitting a more efficient system than requiring each individual to find a source for their information and request it personally. This streamlining of the process facilitates users getting the information they need, but does not benefit the library helping the user, other than that they are serving their population.
It will be decades, if not centuries, before all the content in a research library is available electronically or otherwise through a commercial vendor. In many cases, it might not be profitable for the vendor to provide access to the vast majority of obscure older works. So long as the vast majority of the knowledge of humanity is housed in print volumes, these volumes will become more and more fragile, and loaning them will create ever greater problems. In order to preserve our printed heritage, we need to be able to provide the content in the way that is best for both the present and the future, whatever that method may be.
Even after items are digitized, they may have analog roots. Other items will be 'born digital,' only to be printed out and photocopied. For all concerned, copyright law in this era needs to be open to imminent convergent technologies and to not create artificial and temporary distinctions based on format type. As video games morph into movies and texts come with images of the text encoded on an enclosed CD, distinctions regarding format and delivery method become unworkable.
A.2. A flexible standard is more appropriate and should replace the single-copy restriction. This standard should apply both to copies for a library's in-house users and to copies for ILL.
Incidental copies made in preparation for an ILL transaction are not legally significant. These copies are operational necessities and should be considered part of the process of delivering the 'copy' to the requestor. We urge that copies stored briefly in computer memory and the like should be excluded from the scope of the Copyright Act.
Absent such an approach, Section 108 should be revised to make clear that it is acceptable to provide a digital copy to the requestor, even though one or more incidental copies might have been made in order to deliver the product to the user. As mentioned, none of these incidental copies are kept by the borrowing Library; they merely serve to deliver the material to the user and are not retained by the borrowing Library. We would welcome the clarification of the language of 108(d) and 108(e) to eliminate concern about the implications of incidental copies.
A.3. Interlibrary loan is a fairly common activity that libraries engage in to meet the information needs of their users. Only a member of the requesting library's user community can make an ILL request. As noted above, libraries do not normally make direct copies, even for their own users, because library personnel resources are severely limited.
Libraries generally prefer to make the tools available so that users can make appropriate copies for themselves under Section 108(f). Digital indexing will make it easier to find obscure items in distant libraries, and it is unclear whether most users will go to the effort of requesting access to those materials. Even if users do request those materials, in most cases, it is much more likely that the work in physical form will be sent, because it is easier and less expensive.
We do not believe that digital reproduction and distribution in and of itself will increase the use of digital ILL or other lawful copying. In fact, as immediate access to information via Google and other full-text searches increases, ILL may decrease as users are sated with what is immediately accessible at their fingertips. In contrast, ILL requests generally take a few days to be filled, reducing their benefit to requesters with a short timetable.
This forecast is unlikely to apply to academic law libraries, where intra-library loan may increase as the use of remote storage and the creation of branch campuses increases. Even if there is increased access to digitized information from analog sources, we do not anticipate the number of intra- or interlibrary loan requests related to digital duplication of print materials to drastically increase, even if digital reproduction and delivery were explicitly permitted within Section 108. As the user has already found enough information about a digital copy to develop interest in it, the actual content may already have been digitized, decreasing the need to go back and redigitize the print source.
A.4. Rarely do libraries or archives use subsection (e) for direct copies or for interlibrary loan copies. It is far less expensive, in terms of both staff time and money, to send the requested volume. Permission for digital reproduction and distribution would not increase this already uncommon occurrence.
A.5. Concerns about digital transmission of material from libraries to requestors is already addressed in 108(g)(2), which excludes systematic copying from the scope of the exception. In addition, the ILL requests that involve copying, as opposed to loan of the original, generally involve snippets of the whole, which have little value except to the requester.
Sending digital copies to requesters is more difficult than one might imagine. Such attempts are often thwarted by spam filters, virus filters, incompatible formats, incompatible software, or other challenges. Creating a digital copy from an analog source requires standing at a scanner and copying the section page by page, saving it, then processing the file. Even sending a section of a digital document involves cutting and pasting it, and if possible, converting it to Word or a .pdf file – ensuring that the format is readable by the requester, and then attaching it to an e-mail or uploading it to a server.
A.6. The proposal to limit copying for users and via ILL by formally defining a Library's 'user community' would take a good deal of effort and would provide little in return. Libraries are generally under-funded and understaffed. ILL requests, as well as copying for local users, take staff time, up-to-date technology, and space in the institution, so libraries do not seek to serve users outside their communities in any case. In fact, most law libraries already explicitly restrict the population to whom they provide this labor-intensive and costly service. As consortial arrangements increase, possibly increasing pressures on interlibrary loan within the consortium, there will also be more consortial database licensing, which in turn may cause a decrease in interlibrary loans within the consortium.
A.7. Interlibrary loan transactions in law libraries are by definition mediated, so there is no need to change the law in this area. It may appear that user-initiated requests via an electronic form are unmediated, but in fact these requests are reviewed by ILL staff before processing to ensure that the user is not requesting something that the library already owns. In addition, the borrowing library keeps records when ILLs are filled. Mediation at both ends is needed for the system to work reliably.
A.8. Regarding the distinction between tangible media, such as CDs and DVDs, and intangible media, such as computer files residing on a hard drive, there is little appreciable difference between the two. Does a hard drive suddenly become a physical object when it is removable? How does that change when it resides in a portable laptop?
Even when there is no physical object, the file received by the requester is still a copy, with no additional 106 rights attached. Any subsequent copying still falls under Section 106 and its exceptions, including fair use. Ownership of an electronic file should be treated the same as ownership of a physical object. Borrowing libraries cannot (and do not) presently use photocopies or digital files to expand their collection or archive them on the remote possibility that they could be used to fill future ILL requests, and there is no reason to expect this to start in the future. The current statutory language is format-neutral, whereas the proposed language is digital-specific. Suppose an analog source is scanned and delivered to the library user on CD. It is digital, and it is also physical. In such cases, both limitations might apply, nonsensically. The current law is sufficient.
Materials are often requested through ILL when their utility to the project at hand is undetermined. The importance of Section 108 is that it permits the library to provide the requestor with an item whose utility is later to be determined. Such works are often requested on the basis of a casual recommendation, an appearance in a footnote, or because the work is out of the home library's scope or is out of print. The products of such random requests do not add value to a library's collection, even if the practice were legal, which it clearly is not under the statute as it exists today.
If a particular work is requested through ILL multiple times, the requesting library usually attempts to acquire a copy of the work. As an under-funded service, it is a higher priority to provide the content our users actually request and will use in as timely a manner as possible, and often we will defer more speculative purchases in order to provide an item libraries know is needed. Librarians are scrupulous about adhering to copyright law, to the point where they are sometimes "chilled" into not making a copy that likely would be legal. The concern that copies made in the process of filling a request would be used as a surrogate for library access or ownership may be understandable from a rights holder's point of view, but it has no real basis in fact. Librarians are already well aware that copies provided though ILL do not become part of the borrowing library's collections. Digital copies are not treated differently.
A.9. The fundamental role of libraries is to facilitate access to information. ILL equalizes the field somewhat so that poorer libraries can borrow materials from richer ones. Local public libraries can borrow materials from distant university libraries. Medical libraries can borrow from law libraries. As libraries generally do not borrow materials via ILL that they already own, concerns about ILL displacing creator and distributor income should not eliminate the entire system. There is a middle ground.
One way to avoid this hurdle is to consider the current distinction between 108(d) and 108(e). In the past, books and periodical issues were the smallest available units of content. If services provide end users with article and chapter-level access, we have merely shifted the grain size and the examples in 108(d) and 108(e) can be revised to reflect the newly available content on the basis that substantiality is no longer based on the book or the issue, but on the article and the chapter. We would suggest that 108(e) should be amended to apply in the additional case that the version available does not provide what the requester needs.
Another problem with limiting ILL is that sometimes content is not commercially available in the form in which it originally appeared, or in the form that the requestor requires. For example, a requester may need an article that is available online at a fair price, but if the requester is an law student who needs a color copy of an image for a copyright paper and the available copy is available only in black and white, such a request should be filled via ILL with a color copy, rather than being rejected on the basis that the text of the article is available and should suffice. The same kind of problem occurs if the available digital copy lacks charts, tables, advertisements, or pagination – if such elements are needed by the requester.
Whether or not to provide an ILL item depends on many factors, and librarians – who have long been guardians of copyright protection – should not be prevented from making considered decisions for the benefit of their user communities.
A.10. Although the CONTU guidelines have worked well for many years, they are too rigid for new technological advances that have occurred in the publishing and distribution of copyrighted works. More sensible guidelines could go to the heart of the matter, which is that copies should not be substituted for proper ownership and/or access to a work that is protected by copyright.
However, the CONTU guidelines do not currently have the force of law and should not be 'codified' in the future, but should continue to be recognized as customary law.
A.11. Any revision should remain silent on the subject of international ILL, which should be dealt with through multilateral treaty or some other, more carefully considered method.
C.2. The EU model referenced by the Study Group is too restrictive. Librarians are educators. In teaching a group how to use an unlicensed database, one might display a screen image on a projector for the class. At the very least, if the EU language is adopted, the concept of educational use should be on a par with research use. One reason not to accept the EU solution is our own notion of fair use, which includes display and performance in the phrase, "or by any other means." The key to this topic is to determine what traditional or novel uses might infringe display or performance rights but are not covered by the fair use doctrine. Such cases should be viewed as a whole to determine which uses should be permissible in keeping with the goals of Section 108 and 109.