Donald J. Waters, The Andrew W. Mellon Foundation
Download: PDF Version WORD Version
The curation and communication of information are vital to
the development of scholarship generally and essential to the
health and progress of specific discipline-based teaching and
research. Moreover, just as the accumulation and dissemination
of primary sources of data and secondary sources of interpretation
and analysis serve as the foundation for knowledge building
in the academy, sound principles of information management
and communication are essential for successful decision-making
in a wide range of business, governmental, military, and other
pursuits. Over the last several decades, not only has the growing
application of computers and networking demonstrably improved
the quality, lowered the costs, and speeded up the work of
information curation and communication in many areas, but also
work with digital information has opened entirely new perspectives
and made advances in knowledge possible in the sciences, social
sciences, and humanities that heretofore have been difficult,
even unthinkable. These radical, and unanticipated changes
have led to cogent and persuasive calls for substantial, long-term
programs of investments in research and engineering that would
foster and support the development of “ubiquitous, comprehensive
digital environments” that are “interactive and
functionally complete.” 1
The technical research agenda for creating such a “cyberinfrastructure” is
ambitious. Breakthroughs are needed in a broad range of areas,
including:
- Robotic and other methods for capturing data rapidly,
accurately, in large volumes, and in a variety of formats;
- Facilities
for representing and traversing the multi-lingual and multi-layered
semantic dimensions of those data;
- Simulation, visualization,
and other tools for exploring, comparing, analyzing, and
synthesizing information;
- Collaborative filtering, recommendation
systems, and other innovative techniques that improve and
personalize precision
and recall; and
- Systems for ensuring
persistence of data and software systems.
The purpose of this document, however, is not to justify the
need for and define a program of research to stimulate breakthroughs
in these technical areas. Rather, it is to argue that creating
a national and international cyberinfrastructure will equally
require explicit and ambitious attention to organizational
design and other social issues, which must also be advanced through
a rigorous and thoughtful program of research and development.
The Need
As NSF and the scientific communities contemplate the creation
of a cyberinfrastructure, the need for systematic attention
to questions of organizational design appears in at least three
areas. First, the research agendas for the digital libraries
and related programs have regularly made assumptions about
the organization of various social factors. The RIACS report
on information management, for example, characterizes use,
privacy, security, and usability as “attributes of interoperation,” interpreting
these factors largely as independent variables which define
technology requirements 2. These key social factors, however,
are themselves highly variable, depending at least in part
on the type, purpose, and structure of the organization in
which information and technology users are embedded. Because
so much technical research in this field depends logically
on assumptions about the social organization of information,
the validity of the research—and its potential for further
development and implementation—depends on much more careful
investigation than has been achieved to date of these organizational
assumptions, including those about business models and modes
of operation.
Second, the Atkins report on cyberinfrastructure wisely distinguishes
among the processes of research, development, and operations
as essential for promoting and sustaining the software and
hardware products of an advanced cyberinfrastructure program
(ACP). These three broad classes of activity, of course, are
interrelated and ideally feed back options and requirements
to one another. Such feedback would be an essential component
of the ongoing vitality of the program, but it is not sufficient.
To administer and help sustain ACP, the report further recommends
both an internal organization within NSF, and a community-based
structure of centers. These centers would foster development
activities as well as user support and other operations at
both generic and disciplinary levels. Although the recommendations
for organizational change at NSF are quite specific, the design
of the community-based centers is left unspecified, and will
presumably vary, in part, with the needs of the academic communities
they are intended to serve. However, if the centers are to
help sustain and support the products of ACP, they must be
designed in such a way that they operate in a business-like
fashion and can sustain themselves, eventually independent
of NSF support. Such design should not fail to be informed
by current expert understandings and additional targeted research
that would indicate how certain types of mission, leadership,
governance, organizational structure, legal arrangements for
intellectual property, and financing, especially in the context
of public goods economics, could contribute to—or undermine—the
success of such centers.
Third, many digital library and related research projects
depend on the use or creation of substantial databases of content
from one or more subject domains. Moreover, subject-based research
within specific academic disciplines also often yields significant
collections of data and other content that may themselves be
valuable outcomes of the funded research, and essential ingredients
of an emerging cyberinfrastructure. However, such databases
are often created only with narrow research uses in mind, and
efforts to move them into more broadly based, self-sustaining,
operational use are often stymied. Just as there needs to be
robust, feedback-laden models for software and hardware products
to move from research to development and operations, there
need to be similar models supporting the life cycle of content
in the sciences as well as other domains, especially for those
forms of information that are specific to the digital environment
and for which new types of organizations are needed because
reliance on publishers, libraries and other traditional means
of dissemination do not, or cannot suffice. If it is to endure,
digital content cannot be an afterthought, but must be actively
developed and curated by designated leaders of specific organizations
who are charged with data management, responsible for safeguarding
property, use, and users, and are fully accountable for their
actions in a clearly defined governance structure. In other
words, just as the design of the ACP centers would need to
be informed by current expert understandings and additional
targeted research regarding organizational factors such as
mission, leadership, governance, organizational structure,
legal arrangements for intellectual property, and financing,
so too would the design of organizations responsible for cyberinfrastructure
content.
A Suggested Program
A hugely fertile area for research and development to support
the curation and communication of scholarly information in
the sciences and other disciplines thus falls under the broad
topic of organizational design. Electronic resources do not
need to be managed within existing organizational structures,
but to persist they must be managed within some organizational
context, and as the previous section has demonstrated, the
emerging cyberinfrastructure presents substantial new challenges
in organization and governance. On the one hand, with investment
in technology, barriers to entry for the creation and management
of digital resources can be lower than they are when the storage
of physical items requires large capital investments in physical
objects and buildings to house them, but small institutions
that want to develop, provide, and manage electronic resources
often lack the sophisticated curatorial, legal, financial,
and other organizational skills that are necessary. On the
other hand, the huge economies of scale that are possible with
digital databases are difficult to manage over current institutional
boundaries. Clearly, new organizations and organizational models
are needed that are sensitive to the dynamics of particular
scientific communities, driven by academic mission, and able
to sustain themselves over time as integral parts of the broader
cyberinfrastructure. To foster the development of appropriate
organizations and organizational models, NSF should institute
the following programs and features as part of the broader
Advanced Cyberinfrastructure Program (ACP):
1) Research on organizations and organizational design. There
would be two broad objectives to this research. On the one
hand, it would identify the organizational variation within
an academic community or set of academic communities that would
affect the requirements and parameters for research, development,
and operation of new technology funded as part of the ACP.
A second objective would be to take various scenarios of research,
development, and operation, and explore the advantages and
disadvantages for the emerging cyberinfrastructure of different
mixes of organizational features.
The research on organizational design should focus on the
following organizational variables: types of mission such as
commercial and non-profit; types of governance, including membership,
board, and partnership models; leadership qualities; structural
dimensions, such as size; policy issues, such as privacy, security,
and risk management approaches to the ownership and use of
intellectual property; and financing options, taking account
of the importance of common or public good economics for the
emerging cyberinfrastructure. Funded work could employ a mix
of empirical case studies and theoretical approaches, and it
could be embedded as part of a larger project or conducted
as a standalone initiative.
2) An apparatus for incubating and supporting new organizations.
In order to create a sustainable cyberinfrastructure, the
new centers and content-management organizations should have
access
to a highly specialized organization—or set of organizations—that
can provide expert advice on questions of mission, leadership,
governance, and general business practices, so that when created,
new organizations created or struggling to survive in the cyberinfrastructure
have a reasonable chance of operating in a business-like fashion.
The supporting organization(s) must not operate in a “cookie-cutter” fashion,
but must be sensitive to variations in need among academic
communities, as well as to differences in size and trajectories
of growth. Ideally, to economize on the costly duplication
of services, the supporting organization(s) might also
take direct responsibility for providing a set of common
services,
such as accounting, human resources, board governance,
and legal advice, thereby helping to create a family (or
families)
of efficiently run organizations.
|
|
