The committee’s motivation was to inspire tool development and
leverage the efforts of computer scientists. Its goal was to "develop a more explicit
partnership between scientists in general and computer scientists in particular" (1).
It was recognized that science’s traditional culture of competition and
individuality, and the traditional system of rewards and recognition among scientists,
were major impediments to achieving effective collaboration. Bottom-up motivation was
recognized as an essential success factor (2). Education and research were given equal
priority (3). A resolution of the perception of technology and tools development as being
somehow "less than" research needed to be reached (4) in order to achieve
an environment in which a scientist’s instruments and information are virtually
local, regardless of their actual locations... [and the] ...virtual presence of an
individual in someone else’s laboratory [can be achieved]. (2)
The 105-page National Collaboratories includes an Executive Summary, six
sections, and four appendixes. Three of the sections focus on specific requirements for
building collaboratories for genome research, oceanography, and space physics. Together,
these criteria provide the generic model for the collaboratory. The final section of National
Collaboratories focuses on the logistics for implementation of a national program.
"Building and Using Collaboratories"
Section Five of National Collaboratories, "Building and Using
Collaboratories" distills the discipline-specific needs identified in its preceding
chapters. It enumerates the basic capabilities a generic collaboratory should support, and
analyzes the social, organizational, technical, and practical considerations for
individuals and institutions.
There are four general classes of information-related problems and specific needs
within those areas that the collaboratory must address technically for large-scale
functionality to achieved:
1. Data Sharing
Electronic libraries combining databases, literature, and software
Easily accessible data archives
A comprehensive system supporting retrieval of data from any and all sources, regardless
of origin or location, and the ability to record logical associations between items in
this federated data
2. Software Sharing
Between on-site data and off-site software
Between off-site data and on-site software
Network-accessible storage of results
3. Controlling Remote Instruments
Reliable networking with sufficient speed to support interactive responses
Method for capturing and transmitting output
Standardized telemetry protocols
4. Communicating with remote colleagues
Including voice, video, text, data, images
Synchronous and asynchronous modes. (56)
Functions between and among these information problems must be transparent (62), i.e.
they must use a single set of commands. Knowledge Robots, or intelligent agents, might be
developed to achieve this transparency.
Social and Institutional Factors
National Collaboratories identifies social and institutional factors relevant to
the individual scientist that must be tackled:
- Motivation for participation
- Confidence and security in data
- Rewards and recognition for contributions
- Resolution between the perceived value of tools and technology construction, and
research (64).
The NRC workgroup recognized that a sense of community was necessary in order to
establish and sustain collaborative communication among scientists, and recommended
starting with scientific communities predisposed to communication while research
developed. The workgroup recognized that the costs of using computer-based collaboration
technology for individual scientists include:
- Incompatibility/critical mass of tasks and people (having to switch between media)
- Economic costs
- Dependence on network and vulnerability (of data and functionality)
- Need for ongoing education and training
- Development of effective, equilateral partnerships (65).
Issues for institutions supporting collaborating scientists include:
- Providing local infrastructure support (including new administrative, control, and
accountability models, and increased support for technicians and support personnel)
- Managing the results of increased interaction, including increased travel and managing
shared results (68).
Structural Issues
Structural issues were the major funding concern for the collaboratory. Since no one
agency is charged with infrastructure development, it is generally perceived that
allocating funds implies taking money away from individual or domain-specific research
grants (70). Fear of funding diversion coupled with the problems that many scientists have
in collaborating in any circumstances led the NRC to recommend that collaboratory projects
be selected from the bottom up,
that is, they will be launched in response to inquiries and efforts by groups of
scientists who recognize a need to collaborate and who manifest an interest in applying
more and better information technology. (70)
Technological Base
The technological base for the generic model of the collaboratory includes five
criteria:
- Interoperability
- Transparency
- Customizability
- Integrity
- Extensibility (28-29)
The Generic Model of the Collaboratory
Robbin (1995, 40) synthesized the collaboratory environment described in National
Collaboratories, and constructed a generic evaluative model of the collaboratory for
evaluation of her collaboratory project. The NRC-Robbins generic model includes:
- A distributed computer system
- Networked laboratory instruments and data-gathering platforms
- Tools to enable a variety of collaborative activities
- Financial and human resources for maintaining, evolving, coordinating and assisting in
the use of computer-based facilities
- Digital archives and libraries that include tools for organizing, describing and
managing data, including images, to enable large-scale sharing of data
- Digital libraries which include tools for organizing, describing, and managing the
information derived from analysis of the data.
The CIRAL Matrix
Phase Two of this study seeks a subjective reality of the collaboratory based on
immersion in the online environment. The subjective reality seeks to determine whether,
and to what extent, the collaboratory actually exists, and whether the subjective
experience corresponds with the objective reality determined in Phase One. To establish
preliminary criteria for evaluating the existence of the collaboratory, Robbin's NRC-based
generic model for evaluating the collaboratory is synthesized and reconstructed. The
fundamental components of the NRC-Robbins model are: distributed Computing,
networked Instruments and tools, support Resources, data Archives,
and digital Libraries, or CIRAL.
The CIRAL matrix (rendered as Appendix B) considers the needs of expansive
research into the other key constituents of the information ecology (people, practices,
and values), as well as the need to explore issues and factors at both the individual and
institutional levels as identified in National Collaboratories. But, to serve the
scope of this study, that is, a synoptic exploration of the environment, the CIRAL
matrix is modified to concentrate on the presence of the five criteria for inclusion, and
to answer the fundamental question: Does the collaboratory exist?
Discussion of Criteria for Inclusion
For a collaboratory to exist, the presence of computerized networks, and by necessity
the resources to support them, are a given. Without computerized networks and the
resources to support them, no collaboratory could be online and so would not exist.
Several articles from the collaboratory literature offer online addresses for
collaboratories. Finholt's (1997) general article in Psychological Science 8(1),
provides a compilation of thirteen "existing collaboratories" (31), and includes
most of the collaboratories that are given individual attention in or are otherwise
mentioned by the remaining literature. Simple searches for the single word
"collaboratory" on major World Wide Web search sites (Alta Vista, Hotbot,
Excite, Yahoo, etc.) produce thousands of links to sites that call themselves
"collaboratory." These preliminary results advance the assumption that
computerized networks and the resources to support them exist, and that there is some
manifestation of the collaboratory online.
Data archives must also exist if the CIRAL criteria for inclusion are met. Data
archives may simply be the collection of computer files and programs that facilitate USE
of the computerized networks for the collaboratory, and so are also a given if indeed an
online collaboratory exists. But, data archives may be much more than a collection of
enabling software, and, indeed, the archives may be of an entirely new sort of data. They
may include archives of data created by collaboratory participants or instruments during
the course of collaboratory activity, such as with virtual "notebook" systems
that allow the logging of text, images, messages, and other digital information in central
or decentralized systems available to collaboratory participants. The question of data
archives raises the question of digital libraries, and the difference between them, which
will be discussed later. But first, the question of remote instrumentation needs
attention.
It is clear from Wulf (1988), Lederberg and Uncapher (1989) and NRC (1993) that the
intended definition of remote instrumentation is a mechanical device or contrivance which
serves as a tool for specific types of work, is rare or not widely held, and which may be
manipulated by someone remotely located using computerized networks. It may be argued that
the computer network is itself a collaboratory instrument. Thus, interactive online
environments that rely on the instrument of the network (as with virtual reality systems
or interactive web sites), or those that rely on the human mind (as with MUDS and MOOS,
which construct text and object-based interactive online environments), may be included as
collaboratories. While it is easy to imagine and exciting to argue that these online
environments are at least of the same family as the collaboratory, for this study the more
strict definition is adopted. Accordingly, any computerized network supported by resources
that has data archives and digital library resources, but which does not include access to
and the ability to manipulate remotely located instrumentation, is excluded as a
collaboratory.
Likewise, any site that does not include access to and support of the digital library
must be excluded. The obvious distinction between data archives and digital library is the
application of librarianship, a professional activity (Abbott 1988) concerned with the
intellectual work of collecting and disseminating information resources to support
specific human activities or goals. Whether this service is digital, as with a webpage of
relevant links, or is an intelligent agent, (a computer program that searches for,
harvests, formats, and delivers relevant data), or is provided in real time by a human
being, is irrelevant to this study. An exact and rigid definition of "digital
library" is also not necessary since the concept is still very much in its own early
evolutionary stage (Spink 1999). For the purposes of this study, however, digital library
is defined as a specifically built and purposively managed collection of relevant and
pertinent internal and external data and information intended to support the knowledge
activities of collaboratory participants. As such, digital library resources are distinct
from, but may include, data archives, which are defined as files that support the
collaboratory's technological functions and/or record its activities.
Certainly, one collaboratory's data archives may be another collaboratory's digital
library, so further distinction must be made. For the purpose of this study, data archives
are defined as the accumulated and saved records that support the collaboratory's
functional existence, and digital library resources are defined as value-added data, or
information saved, accumulated, and made accessible in service to collaboratory
participants.
Conclusion
Phase One's objective reality uses the library literature to prove as practiced
principles Wulf's assumptions of relative equality of contribution to, and inherent
interdisciplinary of, the collaboratory environment, and theorizes that the collaboratory
is an harmonious, ungendered, intellectual information environment. Chapter Six begins
Phase Two's subjective reality with an investigation of the collaboratory's third
foundation document (NRC 1993), which sets the instrumental foundation for the
collaboratory and from which an evaluative instrument, the CIRAL matrix for
inclusion as a collaboratory, is derived. The necessary but not sufficient criteria for
inclusion as a collaboratory are identified and discussed. Chapter Seven reports the
collaboratory testsite experience, and Chapter Eight presents four descriptive studies as
empirical evidence of the study's immersive phase.
