National Academies, 21st Street and Constitution
Avenue, NW, Washington, DC
January 29-30, 2007
Sponsored by:
Committee for Economic Development
Council on Competitiveness
National Science Foundation
Science Commons
University of Michigan
Selected papers from the conference are available as a special issue of First Monday.
The Internet grew rapidly as a general-purpose enabling infrastructure, upon which controls had to be superimposed in an ad hoc manner. Advanced knowledge infrastructure, by contrast, is human-centered and optimized for particular resources and communities. Cyberinfrastructure offers a vision of advanced knowledge infrastructure for research and education that integrates diverse resources across barriers of geography and time – and across the subtle and complex barriers of discipline, community, sector, and jurisdiction. Because it aspires to provide human-centered access to diversely controlled resources, cyberinfrastructure must be open and sensitive to institutional, legal, and cultural context, especially mechanisms and procedures for collaborative research and innovation. This is critical not only for optimizing the productivity of particular research communities and initiatives but in developing public policy for advancing knowledge and innovation in an IT-enabled world.
Thanks to digitization, enabling technology, globalization, and the rise of collaborative research and innovation, knowledge is generated, organized, and used in a multitude of new ways. Today cyberinfrastructure provides a driving vision for advanced knowledge infrastructure in a complex, vastly expanded, and nearly borderless economy – a customizable, human-centered environment that can be tailored to any task, enterprise, or field of study.
The Internet is the paradigmatic example of enabling infrastructure, founded on nonproprietary technology and minimally regulated. Virtually free of controls, private or public, the Internet spread quickly as a generic technology platform that anyone was free to use and build on. Universal global accessibility has been the default; limits and boundaries had to be engineered in. Controls for service priority, privacy, copyright, and security have been slow to develop, not only they are unsupported in the underlying infrastructure but because they are perceived as departures from the open architecture of the Internet. At the same time, legal regimes for controlling data, information, and knowledge have been enhanced and expanded in response to assertions that stronger protections are needed to encourage greater use and spur investment.
|
Although enabling technologies and legally based controls often appear to work at cross-purposes, in practice voluntary means have emerged to mediate between the two. These include a variety of mechanisms, practices, and institutions that bridge enabling technology and controls prescribed by law, market conditions, community practice, and culture. Some are driven by technical or user communities; others are business or legal responses to perceived problems in innovation and commercialization. (See sidebar.)
These mediating mechanisms have evolved in ways that are unique to information management, software, and information technology. At the same time, they appear to be especially important in advancing these fields. For example, the development of standards is an essential ongoing process in support of emerging markets. Open source software development and distribution is practical because it is Internet-enabled. Open access publication depends on the commodity infrastructure of the World Wide Web. University-based technology transfer has worked very differently for software than for biotechnology because much software is non-resource-intensive, targeted to academic use, or supported by users.
Information technology and infrastructure have a powerful two-way enabling relationship with collaborative innovation, especially for fields in data-intensive science and technology. Each expands the value and potential of the other. Cyberinfrastructure offers a rich, multidimensional infrastructure extending up the stack and across all fields of science and technology – accompanied by an expanding repertoire of practices and strategies for overcoming the barriers to collaboration between individual researchers, research teams, disciplines, institutions, firms, and sectors.
Building on the Internet
Both the NSFNET of the 1980s and today’s vision of cyberinfrastructure were motivated by the need to share costly high-end computing resources efficiently. In the case of the NSFNET, the need for low-end sharing and communications among a growing universe of users became the driving force behind privatization and commercialization. Although high-end needs still drive advanced connectivity and collaboration technology, cyberinfrastructure trades the “high performance computing and communications” vision of the NSFNET for a vision of human-centered, software-integrated knowledge infrastructure.
As information infrastructure has evolved from standalone databases to email to the multifunctional Internet to a content-linked, transaction-based World Wide Web, expanding capabilities have elicited fresh concerns and sometimes new controls. In the late 1970s there was an outcry over databases containing private and sensitive information, all of which might easily escape the location and jurisdiction where it was collected. This led to the OECD Guidelines on the Protection of Privacy and Transborder Flows of Personal Data (1980), yet similar concerns arose in the context of distributed electronic commerce 20 years later. The functionally sophisticated and ubiquitous Internet also presented challenges for copyright, so new controls were added in the Digital Millennium Copyright Act.
Cyberinfrastructure inherits all of the legal and policy challenges of the Internet and the Web – and adds to them. The challenges initially appear more manageable because of the relatively cloistered, noncommercial nature of scientific research. However, the cyberinfrastructure vision extends explicitly to education and underrepresented groups, and to practical implementation of collaboratories, virtual organizations, and human-centered infrastructure. What was styled in early 1990s as the National Research and Education Network emerged as open and ubiquitous – by no means limited to research and education. As long as cyberinfrastructure, too, remains open to the world, it confronts fundamental issues of access, ownership, control, and consent.
The ultimate success of the Internet and browser technology, both of which were supported by NSF, lay in their widespread commercialization, including the immense variety of proprietary applications and services they spawned. It is increasingly expected that the results of publicly funded research should lead to further development, application, and commercialization. Today, these economic payoffs provide a strong argument for increased public investment in information technology and infrastructure.
The Internet spawned the development of a vast array of proprietary products and services, in part because the basic technology was fully nonproprietary, and anyone could build on it without seeking permission. However, the proliferation of software patents and broad “business method” patents (including patents on tools and infrastructure for education and research) has led to a mixed environment, in which it may be difficult to determine what permission is needed and from whom. At the same time, mechanisms to preserve and pool knowledge as a common resource have emerged, sometimes in reaction to perceived problems of overpropertization and proprietary control.
The report of the National Innovation Initiative frames the problem and the opportunity for accommodation:
[O]pen and proprietary IP models should not be seen as mutually exclusive; rather, the IP framework must enable both approaches. Because collaborative innovation is relatively new, however, the structure and processes to accommodate ownership, openness and access are evolving. New creative models are emerging across sectors. (Innovate America, Council on Competitiveness, December 2004, p. 44)
In collaborative environments with few participants, there may be little need for controls, let alone mediating mechanisms. Shared norms and a few readily understood rules may do the job. However, as environments grow in scale, scope, and significance, they naturally become more heterogeneous, less governable, and less secure. More capabilities are demanded, but common understandings become harder to express and enforce as barriers to the outside fall, and the rules of the market and public law take precedence.
Cyberinfrastructure must grow and mature at a level and time when controls are stronger and more pervasive, more divisive, and more controversial. Deep industry differences have stalled patent reform legislation in the U.S. The U.S. and Europe have taken conspicuously different approaches to control of personal data, ownership of databases, the scope of patentable subject matter, and the research exemption to patent infringement. These differences persist despite common enabling infrastructure, common concern for promoting innovation, and a shared commitment to international harmonization.
Yet as cyberinfrastructure expands, it can draw on a growing menu of mechanisms, institutions, and strategies for reconciling enabling infrastructure and rights-based controls. Optimizing these tools to support research and innovation requires better understanding the social and economic forces surrounding the processes of science and technology. It requires a disinterested, empirically grounded perspective on how tensions between enablement and control are addressed and mitigated in practice. Experience with expanded infrastructure should tell us much about the ecology of knowledge and innovation in a world of porous boundaries and immensely powerful tools. This, in turn, may help us rethink and redefine laws and policies inherited from simpler times and circumstances.
For further information contact Brian Kahin, kahin@umich.edu.