The future of the library is a topic of great importance for the higher-education sector as well as for the library sector. I bring a university perspective to the topic. My focus is on developments in higher education and the influence of these developments on university libraries. I assume that university libraries will adapt to change in education and research institutions as they are transformed through the digital revolution. Many libraries, whether in higher-education or not, are strengthening their education missions in their own right. However my focus today is on universities, on the need for change in their information services, and on the broader range of services which university libraries might provide in meeting that need.
The Library Enterprise Faces the Future
Emerging visions-the library of the future, or the future of the library-is a topic without boundaries. Two key questions have focused my thinking. First, what futures are enabled for the library by the remarkable development of digitized text, visual, and audio materials and by the equally remarkable development of global, undifferentiated, end-to-end digital communications? Second, what are, for want of a better term, the “business models” for one future rather than another? By business models, I mean to include desirable objectives, strategies, and-most important-feasible funding arrangements.
The library enterprise, construed broadly as services that preserve and make available published and valued information, is coextensive with what we have come to call the global civic information infrastructure. But this infrastructure and the library are quite different enterprises. My comments will focus on a third enterprise, the academic enterprise, a venture largely prescribed by education, research, and research training. I keep these three enterprises-the academic, the library, and the information infrastructure-distinct in my mind, even though they are often conflated when the impact of technology on higher education is discussed. It is easy to see why they are often intermixed, especially when the focus is on libraries and universities, given that universities institutionalize pursuit of the academic enterprise. But in so doing, universities also become major contributors both to the library and to information infrastructure enterprises. It is in our interest to understand these entanglements, to separate concerns, and to be as deliberative as we can about the information services that support the academic enterprise, whether they be sourced in library agendas or elsewhere.
University Objectives for Digital Assets
Because institutions can be conservative to a fault-in the university sector at least-we have yet to rebuild resource-allocation processes and to take into account the value of the new generation of support services for the academic enterprise, otherwise so tantalizingly close at hand. Many of our problems arise from not having reviewed and restructured the way in which we manage our resources. Not surprisingly, consideration of business cases for servicing the academic enterprise quickly leads to consideration of the role of institutions and how they behave. Institutions can play a vital role in creating libraries of the future, and it is a role that they are destined to play. Sometime soon, many universities will adopt the following objectives in one form or the other:
- They will manage, preserve, and provide open access to their digital assets. Managing digital assets internally includes managing rights for others to use these assets. We are already involved in managing information on the corporate side of universities through content management systems, including metadata issues that are similar to, and in fact overlap, metadata challenges associated with scholarly information.
- Universities will preserve digital assets that are judged to have long-term value; how they make those judgments will become clearer with time.
- Universities will provide open access to their digital assets, including elevation of these assets into global access platforms; develop digital asset holdings in line with their strategic interests; and foster and sponsor national and global communities that will be built around education, research, and research training.
None of these objectives is new, but each of them needs to be qualified by the extent to which they will-or will not-strengthen individual institutions. I believe that adopting these objectives will strengthen most institutions, and that such action, in turn, will create the need for services that are quintessentially associated with librarianship. National interests sometimes surface in this agenda; however, the drivers are international in character.
What issues are raised for our institutions in pursuing these objectives? It is clear that new services are involved and that those services will require skills that are not readily found in our current service base, including libraries. I expect librarian-based services will emerge that embody those skills. Further, while library facilities will continue to have a strong physical presence for campus-based education programs, the services we are thinking about must be available when and where they are needed, and many of them will be projected through the information infrastructure.
Moving toward Communities of Interest
Because significant shifts in funding for services are unlikely, introducing new services will mean displacing existing ones. This displacement is likely to be difficult if we do not build programs for staff development-for academics and students as well as technical support staff-to take us there. We have to get our heads out a few years, see where we are going, and make sure we can go there with our values intact. We also must make sure that these values come from the broad community.
To meet these objectives, academics will need to become engaged in information management processes. Discipline-specific judgments are required. The paramount need, as I see it, is to build academic practices while introducing new services. This combined approach argues for “communities of interest,” a notion to which I assign an explicit place in the strategies for moving ahead. In the sense I use it here, “communities of interest” are broadly defined as the academic communities that have common or cognate research methods inasmuch as those methods depend on particular information services. The distinctive needs of musicologists and visual anthropologists in working with sound and imagery, respectively, are examples of such communities.
Digital technologies are the drivers for change in library services. They are disruptive; otherwise, we probably would not be here talking about these issues, and they are arguably the most enabling of all the technologies shaping our social and economic systems. Change over the past decade is astounding, and the pace of change continues to increase. Possibilities for the future appear boundless. Not so long ago, the call for digitization programs was a familiar one. The vision was simple: Encode our information sources and digital databases, either by conversion or digital creation, and much of the promise of the information revolution will follow or, at the least, become possible.
We are still at an early stage of converting existing information sources to digitized forms. The world, however, now has a burgeoning layer of digital data, much of it contained in the global information infrastructure. In part because of the vast quantities and ongoing proliferation of data generated, we are developing a sharp focus on the need to understand how data can be structured, valued, integrated, preserved, and accessed through purpose-driven discovery processes. The purposes here are the higher-education purposes.
Validation, Creation, Discovery, and Preservation: Prescriptions for the Institutional Community
These challenging dimensions of data management become considerably more complicated when the rights of creators and owners are taken into account. Higher-education institutions have a special interest in ensuring these challenges are met, especially for information whose authority stems from scholarship. Scholarly information is validated within academic communities. That point is made frequently and in a number of ways, but this concept needs to be kept in mind: By and large, we do validate our own information sources.
Creativity is deeply respected in academic communities, and the traditional open publication of scholarly works has provided ready access to ideas for the wider community. Boyer’s classification of scholarly activities-that is, his four scholarships of discovery, teaching, integration, and application-helps us appreciate the broad range of materials involved in scholarly activities and to which we assign high value in-house. This range has expanded dramatically. Traditional papers and manuscripts are now accompanied by data sets, programs, compositions in audio and visual domains, and a wide range of multimedia objects. The extent of this expansion and the variety of forms scholarly works take are straining the traditional peer-review processes on which validation of materials rests.
Indeed, the meaning of the term “scholarly publication” has been set adrift. Universities build, nurture, and foster scholarly communities. They secure and manage financial resources and provide physical and informational infrastructure in support of those communities. The wherewithal to create, collect, store, discover, organize, preserve, and access scholarly materials is a vital part of a university’s support. Is there a boundary between the academic and the library enterprises? I do not know the answer to that question. In the past, there has not been too much of a division, and I hope that there will not be a boundary in the future.
Creation of scholarly materials excepted (admittedly a big exception), responsibility for information management has largely been institutionalized through facilities and services set up in commons, with libraries having the frontline responsibility. Within this model, all the aspects of information management other than creating-in other words, collecting, storing, discovering, organizing, preserving, and giving access-are going to be changed dramatically by digitization. That change is happening already. Our institutional structures, however, will also change in the divide between the commons and the rest of the university, and that divide is very important.
The evolution of skills and expertise underpinning traditional library services has been strongly conditioned by print technologies and the relatively shallow, largely discipline-independent information structures upon which organization and discovery of information have been based. This is, of course, the case more for published materials than it is for primary material collections, the latter usually involving a deeper understanding of content. Primary materials or not, support for organization and discovery of print objects at the infrastructure level has been built around catalog structures and strongly encapsulated print objects.
The Consequences of Digitization for the Academy
I want to consider three consequences of the emerging digital framework for the academy. First, digital objects may not be strongly encapsulated. They are inherently compound objects, having internal structures that can readily be exported to the discovery-organizational-access processes. Given sophisticated indexing mechanisms, which we do not yet have, parts of a digital object can be reused. Boundaries among objects in broad collections are blurred by a sharing of subcomponents. These boundaries can almost disappear or be multiply determined according to specific purposes to the queries, ownership, or rights restrictions.
A second consequence is that academic processes are increasingly being conducted in ways that depend on digital-to-digital processes. This dependence, in turn, leads to incremental publishing and to the need to consider the information structures within and between digital objects early in research and education programs, rather than in the dissemination phase.
Third, publication of the wide range of digital objects associated with education and scholarly research leads to the need to attach comprehensive descriptions of great integrity to individual objects and especially to objects that are not otherwise self-defining. I am thinking here well beyond the normal metadata discussions and considering the emerging ontological dimensions.
With these three consequences of digitization in mind, it is not difficult to envisage the new services that we need to introduce to serve academics better in this digital world. Will librarianship embrace the information professional skills that are being pointed to here? I hope so. I expect so. In any case, broadening the skills base to include the design and specification of the information structures is an obvious need, as are the skills to define access and rights-management structures.
It is less obvious how to address the need to have information professionals contribute to the design in the early stages of research and education programs. The viability of university-wide services is weakened as direct involvement in individual programs is increased. On the other hand, it is not practical to fund all these programs as a single entity, including having information professionals as team members of all projects as a matter of course. This has led me to focus on building communities of interest. This approach is not revolutionary, but it does invite a focus that I think we should strengthen. More important, communities of interest provide an approach to resolving the tension between individual programs and university-wide services.
Australian National University: One Institutional Approach
It is difficult to prescribe the impact of information technologies on universities. However, we can aim to create responsive institutions and learn from one another as to how we might do so. For this section of my talk I have drawn on the pathway the Australian National University (ANU) has chosen. University organization structures per se are not particularly relevant to the cause; accordingly, my comments will refer to functional aspects of the ANU’s experiences.
The ANU promotes a three-component functional model which includes platforms, communities, and services, and that focuses on bilateral relationships among the three components. Platforms are the enterprise-level systems-repositories and operating processes, and their links to the national and global information infrastructure-used to support academic communities. The service component comprises skills and expertise that, on the one hand, assist communities in using the platforms and, on the other, assist in developing the platforms to better support research methods and outcomes associated with particular communities. The expertise and skills needed in the services component guide us to the expanded roles of information professionals needed for the university library of the future.
The ANU’s first move in responding to the impact of information technologies occurred in 1999-2000, when it carried out a major review of information policies-a process that most universities have performed over the past decade. Our institutional goals for the academic enterprise were used to drive objectives for our information infrastructure; again, a quite common approach. We then looked at support services, not only scholarly services but also corporate services. All universities involve people working in multiple roles-research, teaching, learning, and administration. We tracked the major roles in which any individual was involved and then mapped out the dependency between scholarly and corporate information services. The trends showed us that teaching and learning-even research, to some extent-are increasingly dependent on corporate as well as scholarly information. Accordingly, we set about coordinating these previously separate areas to provide a broad set of academic services supporting both classes of information.
We also recognized the need to build staff training and information literacy programs. We have since put such programs in place; however, they are resource-intensive and will require restructuring within service budgets if they are to be sustained. The main outcome, at least from the perspectives of governance, organization, and coordination, was to create an information portfolio to sit alongside the portfolios for research and education. Under this arrangement, university strategy and policy formation involve information agendas as one of three major portfolios.
To complement portfolio arrangements, the university created an Information Strategy Committee as a subcommittee of the Academic Board. Then to run strongly coordinated programs across campus, a converged, single budget, Division of Information was established. The division is responsible for converged services covering four areas: scholarly information (including libraries); corporate information (including enterprise systems); scholarly technologies; and information technology infrastructure. Coordination aside, the main purpose of putting these areas together under the Division of Information and giving them a one-line budget was to balance resource allocation and priority setting.
A Planning and Strategy Framework for the Information Infrastructure
For planning purposes, the academic enterprise meets the information infrastructure under, for want of better names, e-research and e-education headings. This division reflects major funding lines and strategies for many universities. Figure 1 illustrates this structure.
Fig. 1. Planning and Strategy Framework for the Information Infrastructure
The level below research and education reflects divides that, although contingent in nature, are driving infrastructure development in many universities. At the ANU, services in support of these two broadly defined communities currently involve different skills and have different priorities for developing platforms. In any case, communities of interest within these two major headings are emerging, and are engaging the information infrastructure more deeply in carrying out their research.
In broad terms, the science community is at a more advanced level of digitization, and is more comfortable with its scholarly communications being mediated by networked repositories, than is the humanities community. The difference can be quite stark, with science academics expressing delight with change in areas where humanities academics express deep concern. In part, this difference has developed because the sciences have been able to invest more resources in digitization than have their colleagues in the humanities. However, differences run deeper than this. The importance of computational laboratories, sharing of research data sets, and time to publication have moved infrastructure toward the sciences. We have developed high-performance computational laboratories and large-scale storage facilities linked by broadband networks. Data grids are springing up across the world and access-grid nodes are now fairly common at research-intensive universities. A thorough exposition of the role of infrastructure for major scientific ventures in the future is provided in Atkinson et al. 2003.
On the humanities side, the same future is in prospect, but it just lags a bit behind. New-Model Scholarship: How Will It Survive?, a recent report by Abby Smith of the Council on Library and Information Resources, provides an overview of the emerging strength of the humanities in the digitized world.
Learning materials are a major focus on the education side. From an institutional services point of view, it is a difficult area because of the cost of producing quality materials. Although the publishing industry is energetically exploring business strategies, materials currently being injected into learning environments include a wide range of institutionally produced documents and related digital objects. In addition, students locate and retrieve materials from the global information infrastructure using generalized discovery mechanisms rather than local library catalogs, even though the latter are online. This experience is widespread in education institutions. Learning environments, most based on Blackboard or WebCT, have become enterprise systems in their own right and are typically integrated with companion staff and student enterprise systems.
When one looks across the service needs to support each of the intersections in the foregoing diagram, the wide range of skills and expertise required in the service component is apparent. It presents a challenge for information professionals in universities, be they librarians, educational technologists, multimedia producers, or systems analysts.
Priorities for an Evolving Infrastructure
I often talk with my colleagues about evolving our infrastructure and about the best information environment they can imagine. In the e-sciences, for example, the major services people want are high-performance computing, collaborative visualization, cooperative environments, information access, and online instruments. At least three of these service categories-collaborative visualization, cooperative environments, and information access-are now being integrated into traditional library-based services. In the process, information professionals who can work across the broader spectrum are emerging.
On the e-humanities side, content creation and resource discovery matter most. The emergence of content management systems is viewed enthusiastically, and free-ranging discussions about the strength of text encoding and XML are commonplace. Other priority areas for institutional support are resource description schema, metadata encoding, learning management, library systems integration, and digital preservation. These topics are targets for services that we need to put in place. Preservation is, of course, a difficult problem. If we look out far enough, we hope that it will not be as hard to in fact achieve a reasonable degree of preservation as it appears to be from the perspective of the wide range of currently unsolved problems. The issue is clearly underrepresented in the multitude of “one-off”-funded digitization projects across the world. Under the resource discovery heading, questions are commonly raised about persistent identification. There are different views on this topic, as between persistent URLs or other, more content-indexed approaches. As in many other areas, awareness of the issues is more important than specific solutions. The ability to search, both as full-object and metadata, is also important, and a difficult problem when material is locked up at the repository level or within the structure of digital objects. The goal is to build a global infrastructure that enables discovery across all interconnected repositories, subject to rights and permissions. Interoperability though the Open Archival Initiative System is a promising framework for realizing this objective.
Progress on the Pathway toward a Cohesive Information Infrastructure
At what point has ANU, as one higher-education institution, arrived? If we had not reorganized, we would not have moved nearly as far as we have along the pathway toward incorporating what we see as future library services. It is appropriate to emphasize the importance of skills development in this journey. The ANU has been quite aggressive with information literacy programs, allocating substantial resources and paying close attention to quality and participants’ needs. Services will not be effective unless providers and receivers have a shared knowledge base and common expectations.
Regarding digital asset management, universities have a clear need for content management systems for both corporate and academic information. Although common platforms are not practical in many institutions, it is practical to run programs that keep metadata aligned, and there are substantial benefits in doing so. Toward open access to digital assets, repository platforms such as e-prints and Dspace are being deployed across the higher-education sector. Rights management is an open-ended problem, and systematic solutions to it are not yet in sight. Although comprehensive solutions are some way off, much can be achieved in scholarly areas through frameworks such as Creative Commons.
Preserving digital assets requires looking at those assets that are judged to have a long-term value. Judged by whom? The university and disciplinary societies can judge, advise, and influence resource allocation to preservation agendas. However, more generally, there is a great deal of archive-worthy content for which there is no effective equivalent of the heritage movements found in physical, rather than informational, domains. This preservation problem has taken its place among the major concerns associated with an information-storage fabric based on interoperating institutional repositories. Although we are fostering communities of interest and developing valued digital assets at an institutional level, we have yet to quantify the long-term costs.
Barriers to Communities of Interest
There are a number of barriers to building communities of interest with respect to information infrastructure services. A major difficulty arises because ways of valuing digital objects as scholarly “productions” have yet to emerge. Even within the academic communities that have produced them, the value of digital objects can be elusive. The question of how to value the scholarship that goes into digital object design and implementation is also fundamental.
Another challenge comes with elevating objects into international and global access platforms. We can engineer such platforms, but technologies and associated standards are at a relatively early stage. Nonetheless, the development of global access platforms is widely seen as just a matter of time, and there is a great deal of optimism that it will be “sooner rather than later,” at least for public domain information
Rights management presents a problem because of the understanding that the problem will not be solved in the near future. The role that a Creative Commons approach can play has not yet been assimilated into academic culture. As mentioned earlier, preservation of digital assets is also a problem that prevents communities from emerging where they otherwise would. Finally, funding models to sustain a wider range of information professional services is critical to tailoring support for communities of interest.
National and Global Approaches
Australia is energetically developing an e-science infrastructure, informed by, and to a large extent led by, developments in Europe and North America. Through the Australian Partnership for Advanced Communication (APAC), we have a national facility for high-performance computing and associated mass-data stores (see http://www.apac.edu.au). A program for developing a national advanced research and education broadband network was recently launched. In addition, programs based on grid technologies are being planned. Globally, Australia is a partner in the Asia Pacific Advanced Network (APAN).
More generally, a national information infrastructure fund has been set up with a strong focus on scholarly communication. The National Library of Australia is playing a major role, and some of its programs are helping in the higher-education area as well. Australia also has a program, the Learning Federation, concentrating on materials for primary and secondary education.
From an institutional perspective, the value of an e-science infrastructure depends on the extent to which the academic community can use it to strengthen their research. This, in turn, depends on institutional investment in linking infrastructure and, in particular, in enabling services. Grid technologies promise to provide a systematic and integrative approach to the way information repositories, computational modeling, and communications can work together. The information world in this setting is focused on data sets and not yet inhabited by the wide range of professional information skills referred to earlier. However, the demand for such skills is substantial. E-science researchers need to be able to discover where codes are stored and to have access to ontological information.
Large data sets in the e-science world are in general not subject to systematic preservation regimes, and this in itself is a looming problem. A related problem for Australia is to understand, at a technical level, the implications of relatively remote connectivity to the rest of the world. If books are going to talk to themselves, as Marvin Minsky once explained, they had better be network-aware. This issue is another aspect of the skills base needed in campus-based support services.
The Library Enterprise Is a Rich Cultural Story
I have outlined how we need to develop services and an associated skills base for institutions such as universities to manage their own digital assets and to ensure their preservation. There is an overwhelming case for this skills base and a corresponding case for the role of information professionals in the future in higher education. The question asked by this workshop-“What is the vision for a library of the future?”-is answered in large part by this institutional need, at least insofar as the focus is on university libraries. Of course the library enterprise more generally is much broader, deeper, and richer than the university library issues which I have raised. In becoming more closely linked to research and education methods associated with campus-based communities of interest, university libraries will need to embrace a broader set of skills than they have in the past. It is open to universities and their libraries to meet this challenge.
Atkins, Daniel et al. 2003. Revolutionizing Science and Engineering Through Cyberinfrastructure. Report of the National Science Foundation Blue-Ribbon Advisory Panel on Cyberinfrastructure. Available at http://www.communitytechnology.org/nsf_ci_report/.
Smith, Abby. 2003. New-Model Scholarship: How Will it Survive? Washington, D.C.: Council on Library and Information Resources. Available at https://clir.wordpress.clir.org/pubs/abstract/pub114abst.html.
Organizations noted in text
Asia Pacific Advanced Network (APAN) http://apan.net