As new technologies come into being, the world’s cultural objects change shape accordingly; from the clay tablet to papyrus, from the printed book to web-based documents, each takes the form of the technology of its era. Over thousands of years libraries have collected, organized, and made works available (to all, or to a select few) in these formats, and library services have developed to take advantage of the new technologies. In particular, the portability of the printed book in 18th century and beyond meant that libraries could lend works to users, and the mass production of printed texts in the 19th and 20th centuries saw a great proliferation of libraries and the extension of library use and lending to the general population.

The inexpensive reproduction of works has allowed libraries to move their energies from the conservation of objects to the dissemination of highly mobile containers. While the term “library lending” evokes an image of books for most of us, some public libraries in the United States count non-book materials such as music discs, films, and spoken books, as a full thirty percent of the materials they lend. Library lending, however, is both costly and insecure, with both wear and non-returns taking their tolls. Wouldn’t it be great to be able to lend materials that could not be damaged or stolen, and that would be guaranteed to return at the appointed time? This, then, is the promise of digital lending.

Libraries and digital delivery
Libraries have been delivering works in digital formats for over a decade. The delivery of digital works to library users follows two basic models: there is the "all you can eat" model in which users have access to a database of digital materials with no restrictions on how many users can access an item at a time (although licenses may restrict total simultaneous uses to the database from any institution); the other model is an imitation of the lending of hard copy works, and is often called the "one user/one book" model. Within these two models there are different possible delivery options, with some systems presenting portions of materials on the screen but not allowing downloads or offline use, while others do allow downloading of digital items. It is in this latter case where technical enforcement of license terms comes into play, and this is the type of protection that is most often referred to as digital rights management.

The "all you can eat model" is primarily used for research materials, especially journal articles. With the development of large databases of digital full text, academic library users are well-served with instant access to a significant collection of materials. Access to these journal articles is through an institutional subscription, not unlike the subscription to the same materials in paper format. The only technical controls for these materials are on access, which is generally managed through a proxy server on the institution’s network, and which limits access to members of that institution’s community. Users can download and keep copies of articles, somewhat like making a photocopy of articles in the analogue world. The downloaded articles, which are predominantly in Adobe PDF format, have no technical protection that would further restrict copying or printing, although they may be protected against alteration. This model works well for academic materials and will probably continue to do so, although there is some tension between publishers and libraries over costs and over the relationship between the digital license and the hard-copy subscription.

This model is not viable for those materials where units are normally sold individually, especially those materials that might be deemed of a "popular" nature. Books, videos, and musical recordings are in this category. These materials need to use the "one user/one book" model, and require some technical protection on the content files to satisfy publishers that the materials will not be pirated once they have been delivered to end users. In the entertainment arena we have seen the struggle between users and publishers over the unauthorized trading of works in digital form. Books and other lengthy texts have not had the same degree of problems with piracy (for both technical and market reasons), but book publishers have been cautious about delivering their products in a digital form that would open the door to piracy.

The first electronic book products were available only on proprietary hardware, such as the Gemstar (later Rocket) e-book reader. The device protected against unauthorized copying by allowing communication only with the e-book vendor site through phone lines or an Ethernet connection. Some libraries experimented with lending these e-book devices pre-loaded with a selection of books, but the devices did not catch on commercially and the e-books themselves eventually became unavailable.

The first computer-based e-book lending systems that were developed for libraries in the late 1990's, in particular the netLibrary system (cf. sources), required users to read the books online with only one page image downloaded to their computer at a time. This method was used because there was no available technical protection for downloaded files. The books were "checked out" to the library patron and could not be viewed by another library user until the lending period ended. The check out process effectively locked the book so that it could not be accessed until the current loan period ended. Although called "lending," from the user's view this was not at all like using printed books, especially in terms of the quality of the reading experience.

Library lending becomes reality
Although there hasn't been a breakthrough technology that would make electronic reading as popular as its paper counterpart, the availability of software that both facilitates the reading experience and secures the digital content has greatly increased both the willingness of publishers to make their content available and the desire of consumers to purchase that content. Digital content can now by downloaded by consumers to a variety of devices, and can be read off-line.

Libraries have been able to take advantage of the fact that the lending of digital content is compatible with the sale of that same content. In fact, OverDrive (cf. sources), the company whose software is used in book-related e-commerce, is also a major provider of electronic content systems for libraries. In a sense, library lending is the same as a sale, only with a time limit imposed. At the end of that time limit, the rights management software in the downloaded file turns off file access an thus prevents further uses of the content. The book "returns" to the library automatically with no action required on the part of the borrower.

The first lending systems had only one way for the book to return to the virtual shelf, and that was through the expiration date on the loan. This required no communication between the downloaded file and the lending system; each acted independently on the time limit. Even if a user no longer needed the item, it remained checked out and unavailable to others for the duration of the loan period, and because of this libraries were setting very short loan periods, which was discouraging to some users. With current lending technology, users can return a book to the library at any time before the return date. Through an interaction between the checkout system and the rights technology protecting the item on the user's device. This is just one example of how developments in digital rights management (DRM) have made it possible for libraries to provide better service to their users.

Libraries purchase electronic books just like they do their print counterparts through companies that serve the library market. The information about the books is entered into the library catalogue, but instead of a number indicating where the book can be found on the shelf there is a link that takes the user to the virtual shelf of the e-book lending system. All interactions with the e-books go through the library's system, which has user information and authentication routines, and which must record the status of an item ("on shelf," "checked out") for display to library staff and users. Although the user's impression is that the e-book is in the library, in fact the books are stored on a third-party site that delivers the DRM-enabled file to the user's device. At this point in time, the economics of DRM technology do not allow libraries to securely store and deliver electronic files.

Points of purchase for e-books offer consumers a choice of formats corresponding to various brands of reading software and the particular DRM of that brand. Libraries have to select a format when they purchase an e-book. If they wish to have more than one format available they have to purchase each separately, and generally at full price. For this reason, libraries tend to limit their selection to the most widely available software, which today is the Adobe Acrobat format. The Adobe Reader software is available for free for most operating systems, including those of the common hand-held devices which are popular with e-book enthusiasts.

Lending beyond text
Because lending uses technology that is very similar to the technology for sales, in essence any digital formats that can be sold can also be loaned by libraries once the additional lending capabilities are in place. A small number of libraries are beginning to lend audio books. Books "on tape" are very popular items in libraries that lend them, especially in areas where automobile commuting is common. Library lending follows the same model of services as provided by sales points for these files: end users can download the audio book to a personal computer or to a mobile device, or they can burn the audio book onto CDs. All of these actions are secured by the lending system to prevent unauthorized copying of the files to other devices. Although the CD format is unprotected, only uncompressed files are released for these copies. This is the same format that is used in the CD audio books that are sold in stores, and therefore represents a level of risk that publishers have found acceptable.

Lending of musical works and of motion pictures could become technically possible but are not currently available. Some of the issues relate to industry expectations, and others to technology capabilities such as bandwidth. It may also be the case that rights management techniques that are sufficient to protect one form of content will not be suitable for all forms of content. As we see with the relatively low level of protection on academic journals, risks vary both by format and by commercial expectations for different materials. It does appear, though, that the level of rights management that is appropriate for the sale of content is also that which protects the content for library lending.

Libraries: what do they really want?
Lending of e-books and digital audio books by libraries is still very new, and libraries are in the learning stages in terms of what works and what doesn't. From the point of view of libraries, there are some unsolved issues relating to the acquisition and lending of digital materials. These are:

  • Book publishers have a revenue model based on the hard copy world of sales of physical items, but the technology of digital lending does not allow the libraries to actually take possession of the digital item. Libraries must purchase items over which they cannot exercise normal rights of ownership.

  • The storage, delivery, and control of digital materials require sophisticated secure systems. These systems are not affordable to individual libraries, but are usually run as a central service by a vendor. Libraries are dependent on the vendors both for current services and for long term access to materials they own. Should companies fail, and some have, libraries lose access to books they have purchased.

  • There is no one standard format for digital delivery, yet each formatted version of an item requires a separate purchase. At the same time, libraries cannot forego obtaining materials in analogue formats, so increasingly libraries are needing to purchase multiple copies of an item to satisfy the format needs of their clientele.

  • Library services attempt to provide a unified view of the cultural and intellectual sphere, with items from many different publishers and sources treated equally in terms of organization and access. There are many different sources for digital materials, often with their own proprietary technology for access. This may serve the marketing of materials, but it is not conducive to end-user research or bibliographic services.

  • The proprietary formats in which digital materials are issued are not suitable for long-term preservation and access.

Most of these points evidence the difficulties of a transition period in terms of content technology, where the capabilities of the new technology and the market structures in place based on earlier technology are not compatible. The use of individual copies as the basis for the market breaks down in an environment where copies are made each time a user opens a work. One of the promises of digital rights management is that it could re-focus content delivery around rights rather than copies, which could make it possible to solve some of the problems listed above. For example, libraries could be allowed to transform materials to different end-user formats as long as the total number of items in use does not exceed the library's license. The problem of the need for persistent access over time could also be solved by allowing libraries to store a specially formatted archival copy that is not delivered to end users, while at the same time they lend protected copies in consumer formats. All of these capabilities require DRM that guarantees that the digital files will be secure and that publishers will receive payment as agreed.

What this will eventually mean is a move from a market based on copies to a market based on rights. The technology that this will require is not yet in existence, but the required changes are not just technological; huge leaps must be made in the intellectual property markets and in the habits of librarians and those they serve. Some desired features, such as the ability to lend multiple copies when user demand increases for a particular title, are well within the capabilities of the current lending technology but do not meet the accounting needs of publishers, whose system of royalty payments makes the use of micro payments particularly complex. Improvements in the e-commerce middle layer will allow us to experiment with new models of secure file delivery.

Bottom line
Many library professionals view digital rights management as a restriction on use, and it is true that the capability to create restrictive technologies exists. But for libraries to manage and lend published materials in digital formats will necessarily require some controls. If libraries can learn to view digital formats as delivery mechanisms rather than as a substitute for physical copies we may be able to develop a suitable paradigm that is beneficial to libraries and to their users. And if publishers can transition to a revenue model that is based on licenses rather than copies, we will be able to make use of the advantages that digital formats have over their analogue equivalents.


About the author: Karen Coyle is a librarian with almost three decades of experience in digital libraries. She writes and speaks on both technology and policy issues. More information on her work is available at

Status: first posted 26/04/05; licensed under Creative Commons