Digital libraries and scholarly communication: transforming the information chain

Prof. John Mackenzie Owen

University of Amsterdam

owen@hum.uva.nl

The word library has lost its much of appeal during the final decades of the 20th century to the extent that even librarians often prefer to describe themselves as information (or even knowledge) managers. It therefore comes somewhat as a surprise that the word library has become extremely fashionable again. However, it is the word library with a qualifier: the concept everybody seems to be embracing is the digital library. The Journal of the American Society for Information Science recently devoted two entire issues to digital libraries (Chen 2000), the field is acquiring recognition as a respectable area of scholarly research, and there are even serious discussions on the link between digital libraries and electronic commerce, the buzzword of early 21st century economics (Adam 1996).

For librarians whatever they call themselves this development raises an existential question: to what degree are digital libraries related to the concept of a library? In this paper we shall review the relationship between digital libraries and the traditional role of a library, especially in the context of the scholarly information chain. The argument to be put forward is that the real issue is not the transformation of print-based libraries towards new models of digital service provision, but a change in responsibilities for scholarly communication, resulting in an important transformation of institutional roles in the information domain, including the role of the library.

The underlying model for scholarly communication through publications and related documents is the information chain (cf Meadows 1991 and Duff 1997). This is a model which identifies a chain of information functions through which knowledge, recorded in the form of information, is transmitted from the originator to the user, or from producer to consumer. The information chain model identifies specific roles (such as knowledge creation, publishing, archiving, intermediation and use), and actors that perform these roles (such as researchers or research institutes, publishers, libraries and users). Each of these actors performs specific functions within the context of their respective roles. For instance, the functions of publishers can be defined in a task-oriented way (editing, printing, marketing, distribution) or in a more conceptual way (e.g. dissemination, quality control, providing a (canonical) archive, and ensuring recognition of authors, cf. Rowland 1997).

What is interesting about the model is that it not only defines roles, actors and functions, but also implicitly defines responsibilities. In practice, academic institutions regard only knowledge creation as their own responsibility. Publishers long ago have taken on the responsibility for creating information products that serve as a vehicle for knowledge distribution, and libraries are responsible for archiving and availability of information products. There is no overall responsibility for the entire process of scholarly communication, held by a single actor. One might expect academic institutions to bear that responsibility, but until now that has not been the case. As we shall see, that is now changing.

The information chain model has functioned extremely well over the centuries. The various actors have worked together as a co-operative system, and all actors have incorporated new technologies to the benefit of the dissemination of knowledge. However, the model now seems to be at the end of its life cycle. There are a number of indicators that suggest this might be the case. One indicator is that the responsibility of publishers is moving from one based on a contribution to knowledge dissemination - and therefore to scholarship - to one that is based on a contribution to shareholder value. Profit making by publishers has been important for as long as commercial publishing has existed, and it is the basis for the high level of professional publishing we know today. Commercial motives and the advancement of science were always sufficiently balanced to serve both worlds. Increasingly, however, the commercialisation of publishing and the emphasis on shareholder value is creating problems such as the unacceptably high cost of scholarly publications, restrictive copyright practices and insufficient technical innovation. The academic world is beginning to feel that publishers do not serve academic interests in a way that is appropriate in view of their responsibilities and their important role in the information chain.

A second indicator of an oncoming transformation of the information chain model is the evolving relationship between publishers and libraries. The traditional functional distinction between publishers and libraries is becoming blurred as they enter each others domains. Publishers are taking on traditional library functions such as cataloguing & indexing, archiving and end-user services such as document delivery. Libraries are entering the domain of publishing by creating information products such as document repositories and electronic journals. The enabling factors for this are, of course, digitisation and networks, which allow both actors to perform the same functions. At this point it is important to note that technology appears to question the division of functions that is one of the most important characteristics of the traditional information chain. It even begs the question whether, if the distinction between actors is no longer relevant, we need these actors at all.

A third indicator is the large body of initiatives aimed at self-publishing by academics and academic institutions. The most basic form of self publishing is when authors provide access to their publications through a personal or institutional website. But of course this creates a form of grey literature that does not meet the criteria for scholarly publishing recently put forward by Kling and McGim (Kling 1999): publicity, trustworthiness and accessibility. A more structured approach has been developed through the famous pre-print server set up a number of years ago for the high-energy physics domain by Paul Ginsparg. In this approach, authors deposit their work - before official publication - in a database which can be accessed over the network by everyone in the field. Efforts are now under way to group similar initiatives (also known under the name 'open archives') in a consortium to achieve a universal service in the area of scholarly communication (Ginsparg 1999). The major problem with the current forms of self publishing is the lack of quality control through peer review. Possible solutions have been proposed, including up-front payment of publishers to manage peer review, and various forms of 'electronic' peer review over the network (Harnad 1999). The desire of the academic community to transform the traditional information chain and especially the role of commercial publishers - does not only lead to self publishing by individual authors. Especially in the United States many scholarly institutions are now taking on a role as publisher. In this way they try to offer an alternative for the high and ever-increasing costs of commercial scholalry publications, to benefit more from information technology, and to provide some kind of guarantee for archiving and long-term availability of scholarly information. Based on these developments, new models for scholarly communication are being considered (Buck 1999).

Although these various forms of self-publishing by the academic community do not constitute a fundamental change in the underlying information chain model, they do point to a development that might turn out to be crucial for the future of scholarly communication, the information chain and the actors involved. That development is the increasing tendency of the academic word to take on responsibility for the entire process of scholarly communication, rather than leaving the responsibility for specific and important functions to other actors. This tendency has been expresses by David Shulenberger, provost of the University of Kansas, who has urged the academic community to 'devise a collective agenda to address the effective management of intellectual property to protect and promote scholarly communication' and 'to take responsibility for maintaining or creating low cost venues for print or electronic publication of refereed journals of research finding and scholarly thought' (Schulenberger 1998, cf also Alexander 2000).

A fourth indicator, which suggests a more fundamental change in the information chain, is the whole range of initiatives covered by the term 'digital libraries'. The digital library is an often global organization of scientists or scholars who use advanced technology to create and share information over the network. This information can be related to research outcomes, but might also consist of source materials, survey data or data from ongoing research.

The digital library model is no longer based on the traditional information chain, but on a network of researchers who create and distribute knowledge in the form of information objects. These objects increasingly are becoming different from traditional publications such as books and journals; they include text corpora, data collections, audio-visual materials, simulations, embedded software applications, etc. In various digital library projects new document types are being developed, as well as the new technology they require for distribution, long-term archiving, navigation, visualisation and retrieval. One example of these new technologies is the concept of multivalent documents (Phelps 1998), which adds useful functionality to document content.

The digital library model is very different from the traditional information chain. It makes no fundamental difference between author, publisher and library functions, and it relates to a far larger richness of information formats and functionalities. In this model, traditional distinctions such as that between monographs and journals are losing importance: the digital library is based on information objects which could have any type of (often dynamic and distributed) formats. In addition, it is based on an integrated approach to the entire information cycle of information creation, distribution and use, and a new, 'human-centered cyclical model of the scientist's information behaviour (Chien 1997).

The digital library therefore is not a library in the traditional sense. In fact, it is not a library at all, but a new concept for scholarly communication, pointing the way towards a radical transformation of the information chain. What, then, does this mean for traditional actors such as publishers and libraries? Will they become obsolete? Will libraries cease to have a role in scholarly communication other than as an archive or printed materials? Perhaps the main significance of these developments is not that these actors will disappear, but that they themselves will be transformed.

If the academic community expands its responsibility for and control over the dissemination of knowledge, the role of publishers will certainly change. They no longer will be producers of information products, but will act as service providers to the academic community, performing specific functions for which they are paid rather than for information end-products. The key to this lies in existing functions which can be performed adequately by publishers. An example might be quality control through peer review. In any future model peer review will remain important. But peer review is a function that has to be performed by someone. Publishers have an excellent track-record in this area, and there is no reason why they should not be given the opportunity to provide that kind of service, and be paid explicitly for it. The key issue is, however, that publishers will perform such functions in an outsourcing relationship under responsibility of the academic community. That is exactly what a transformation from product-oriented to service-oriented business is about.

Let us now turn to libraries. It is quite clear that the recent history of libraries is characterised by large-scale innovation and modernisation. In Europe, the United States and elsewhere large library R&D-programmes have been carried out, often subsidised by national governments and governmental bodies such as the European Commission. These efforts have produced important results. New systems have been developed which allow libraries to exchange information to provide users better and faster access to a much larger collection of information resources. Libraries have managed to develop many new types of services, notably over the network to the user's desktop. Much also has been achieved in areas such as standardisation, work methods and best practices. New relationships have been established between libraries and between libraries and publishers. The level of knowledge and skills in technical and organizational areas has increased dramatically. Anyone who enters a modern library must admit that it resembles in little the unexciting image of a library that many people still have. For an academic user the most noticeable characteristics are: a generous offering of digital resources, access to the library from any location (including the desktop and home), seamless access to other organisations and - increasingly - also some level of personalisation, i.e. the possibility to adapt the library to one's own needs and preferences.

The technological innovation of libraries has resulted in what can be characterised as a 'hybrid' model of information services. This model combines the two worlds of printed and digital information: some resources are available in digital form over the network, others only in printed form in the library. This model reflects current practice where both printed and digital information are important. However, it is unlikely that this will remain the case. The world of scholarly communication is rapidly becoming fully digital. It is not unlikely that within a few years printed publications will play only a very minor role, at least in a number of scholarly domains. And where there is a need for older - i.e. printed - publications, they often will be converted to digital form. That not only improves the distribution of these resources, but also offers more possibilities for archiving, searching, analysing and processing. Eventually, printed information will become more or less invisible because most users will regard the network as their one and only source of information. This is a practice which already is becoming visible in the information behaviour of many students, and even researchers.

The hybrid form of scholarly communication is firmly based on the traditional model of the information chain and on traditional publication forms such as books and journals. However, the hybrid library appears to be only a transitional phase in a development towards fully virtual scholarly communication model (fig. 1), based exclusively on digital, networked information resources. This is a model that is already becoming visible in the concept of the digital library.

If this view of future developments is correct, then it would seem to be the case that the hybrid libraries as we now know them are approaching the end of their development cycle. That at least is the conclusion of the European Commission, which has now ended its Library Programme and put in its place a new programme directed at multimedia information products and technology. In the Digital Libraries Initiative funded by the National Science Foundation in the USA, large portions of the budget (currently in excess of 100 M$ over the two project phases) have been or are being awarded to projects led by a wide range of academic institutions, not just to libraries.

What does this mean for libraries? Libraries will have to see their future role in the context of a new division of order in the academic world, of which the current digital library initiatives are an early example. What is important to understand, is that the library no longer will hold the exclusive responsibility for scholarly information within academic institutions. That responsibility will be anchored at a higher level in these institutions and will involve more actors than just the library. Researchers and research institutions, in their roles as authors and users, will be in far greater control of everything that has to do with information than libraries are used to at the present moment. This will have implications for funding. Researchers will be the primary recipients of funds for information and projects related to information. The library will have to support many types of information work carried out by others. The librarys contribution will be based on their information skills, not on their information resources. The library will have to expand its focus to many other domains of information, including primary digital sources of research information, and support of information work within research projects. Libraries will have to accept that for many types of academic work the information world is an entirely virtual, digital world which integrates many different and new types of information work and information formats. Development of new managerial and technical skills and creation of new alliances with users will be crucial to the future role of the library.

Libraries are not the driving force in the domain of digital libraries. But they must become involved in that domain to have a future other than as custodians of the printed word.

REFERENCES