CHAPTER TWO
Foundation Documents

The philosophical, intellectual and instrumental foundations of the collaboratory are provided by three key documents. Two of these documents are examined in detail in this chapter. They provide the foundation on which Phase One's analysis rests. Three taxonomies, or quantitative classification schemes, are constructed, and the collaboratory literature (n=86) is analyzed and cross analyzed by discipline, focus, type, topic, and approach of research and publications. The assumptions, practices, and principles of the collaboratory, as put forth in its foundation documents and reflected by the library literature, are explored. The content of a subset of Theory-Type Research publications (n=22) is qualitatively analyzed and an emergent theory of the collaboratory as an ungendered, environmental is put forth.

 

The documents included in the collaboratory literature are those to which access is gained by use of the unadulterated search string "collaboratory." This study does not use truncated or wildcard derivatives of the word. The use of truncated or wildcard derivatives produces a huge field of documents that are closely related, or relevant to the concept of the collaboratory, but not pertinent to this study. Such documents are, for instance, those dealing with the act of collaboration, the collaborative attitude, or people who are collaborators. While these and other concepts are certainly relevant to the development and use of the collaboratory generally, and will be helpful as this agenda progresses, they open a literature base beyond the intent of this study. This study focuses exclusively and sharply on the collaboratory as an information environment. The boundaries of that environment are negotiated as the study progresses.

There is also substantial preceding, succeeding, simultaneous, and derivative literature with strong relational ties to the collaboratory. Those documents are also not included in this study. Barua (1995) provides an analysis of the near meteoric development and evolution of the vocabulary, concepts, and technologies leading to the collaboratory. Access to this substantial literature may be achieved using keywords and phrases including Decision Support Systems (DSS), Computer-based Systems for Cooperative Work (CSCW), Group Support Systems (GSS), Group Decision Support Systems (GDSS), GroupWare Systems, Computer Mediated Communication (CMC), and others. Related concepts may also be accessed using the words and phrases DARPA, ARPA, NREN, World Wide Web, Internet, Digital Library, National Information Infrastructure (NII), Global Information Infrastructure (GII), and their associated concepts.

The intent of Phase One is to focus sharply and exclusively on those documents that are highly pertinent to and specifically address the collaboratory as an environment. The documentary evidence of the collaboratory begins in May 1988, when Wulf coined the word, and, for the purpose of Phase One, ends in December 1998. Pertinent documents obtained outside library research, or outside the time frame of Phase One, are addressed, when appropriate, in Phases Two and Three of this study.

Philosophical and Intellectual Foundations

The documented story of the collaboratory began in March 1989 at an invitational workshop convened by Dr. William A. Wulf, then Director of the National Science Foundation’s (NSF) Directorate for Computer and Information Science and Engineering (CISE). Wulf gathered twenty-nine scientists and researchers to Rockefeller University in New York, and charged them with developing a research agenda to actualize a National Collaboratory.

Wulf developed the notion of the collaboratory in a December 20, 1988 White Paper for the NSF, which he presented to the conference (Wulf 1988, Appendix A in Lederberg and Uncapher 1989). Wulf footnoted in the White Paper that the word collaboratory was "invented to combine the words collaboration and laboratory" (2).

Much in the same way that H.G. Wells' (1938) World Brain visualized a "central intellectual organism" (Rayward 1999), and Vannevar Bush’s (1945) Atlantic Monthly article visualized the memex machine and the hyperdocument environment we now recognize as the World Wide Web, Wulf’s collaboratory vision promised fundamental changes in the way science is conducted. Wulf incorporated access to and remote manipulation of rare and expensive scientific instruments along with interactive human knowledge networks incorporating real-time and document-based communication of various sorts. The collaboratory is a work environment that brings geographically dispersed scientists, instruments, and data together in a simultaneously "live" technologically-enabled environment.

The vision of the collaboratory did not wholly spring from Wulf, however (Banks 1993, Robbin 1995). Yet unnamed, the concept spawned almost simultaneously in several quarters of government and military research in the mid-1980s. Computer and information scientists working on the logistics, languages, architectures, and technicalities of what would eventually emerge as the Internet began to imagine potential uses and use-based environments distinct from, but enabled by the technology. This new science environment demanded separate recognition. Wulf named it the "National Collaboratory."

The National Collaboratory -- White Paper (Wulf 1988)

Since 1988, William A. Wulf has been the AT&T Professor of Engineering and Applied Science at the University of Virginia. Wulf concentrates on undergraduate computer science education, research on computer architecture and computer security, and assisting humanities scholars in the exploitation of information technology. From 1988 to 1990, Wulf served as an Assistant Director of the National Science Foundation, specifically as the Director of the NSF Directorate for Computer and Information Science and Engineering (CISE). In May 1998 he wrote his widely-cited but still largely unavailable White Paper. It was not published nor made publicly available.

In March 1989, Wulf convened a select group of twenty-nine researchers and scientists for an invitational workshop at Rockefeller University, and presumably read or otherwise delivered the White Paper's message. The convention produced Lederberg and Uncapher’s (1989) report, Towards a National Collaboratory, which includes the White Paper as Appendix A. The Lederberg and Uncapher report was also not published. Neither document is catalogued as held by any lending library in the world, nor are they available from the National Science Foundation, or from the authors. Because the White Paper's "center without walls" quote is perhaps the most frequently repeated and often miscited passage within the collaboratory literature, and because neither of the documents was ever published, an in-depth look at both of them is warranted.

The Wulf's White Paper is a two-page, single-spaced document with two sections: "Background" and "The Proposal." In the Background section, Wulf contends that

The health of the United States, economically and militarily, depends on technology... [which] ...depends on the number, quality, and productivity of the nation’s research scientists and engineers... [which] ...depends on such things as adequate facilities, stimulating colleagues, and the open exchange of ideas. (1)

Wulf describes the emergence of geographically situated interdisciplinary centers, institutes, and laboratories which had already produced a "disproportionate share of the advances of their respective fields" (1). He describes the coming trend in which such centers will no longer be geographically determined, but will be "freed from the constraints of distance" producing research teams for which "opportunity and choice will determine the composition, size, and duration" (1).

Wulf warns that it will no longer be necessary for such centers to share a common administrative structure. He foretold a fundamental shift in the way science is conducted. Remote interaction with instruments, colleagues, and data will not only be possible, but mandatory. Interaction with remote instruments will be necessary either because the instruments are too expensive to be widely held (space telescopes), or the environment in which they function is inhospitable to humans (deep ocean vehicles). Interaction with remote colleagues will be mandatory because the talents necessary to address interdisciplinary problems will not be collected in any one place. Remote access to data will be necessary because the data is too vast to be replicated. Finally, some of the most pressing scientific challenges facing us, such as that of the global change, are inherently distributed and exhibit all of these properties; remote interaction with instruments, colleagues, and data is essential to solving them. (1)

Wulf proposed a

major, coordinated program of research and development leading to an electronic ‘collaboratory’, a ‘center without walls’, in which the nation’s researchers can perform their research without regard to geographic location—interacting with colleagues, accessing instrumentation, sharing data and computational resources, accessing information in digital libraries. (1)

The enabling technologies, Wulf wrote, are "high speed information processing and communication" (1). By 1988, a national research network was already underway; high performance computers were becoming ubiquitous; and open system standards were making these facilities more accessible to the research community.

In a sense, the collaboratory was an inevitable outcome of these developments. However, much as a ‘center’ is enabled by the building that houses it but is not just the building, the collaboratory is not just interconnected computers.

A complete infrastructure is required: software that facilitates collaboration, simulation tools that can substitute for some aspects of the traditional wet laboratory, ‘smart instruments’ that can be used effectively remotely and interchangeably [sic] with simulated experiments, digital libraries and software to access the information in them, accessible (usable) repositories of raw data, etc. (1)

A great deal of research still needed to be done "to exploit the enabling technologies and build this infrastructure" (2). Some of the research, Wulf said, was needed in the areas of traditional computer science and computer/communication engineering (network speed, security and integrity of communication, smart instrumentation), but some of the research was essentially social, behavioral, or economic:

How do people collaborate, and how can we exploit technology to amplify the effectiveness of this collaboration, especially when the collaborators are not colocated? (2)

Wulf explained that while much of this research was underway, it was not coordinated and hence was not easily combined; that

when viewed independently, aspects crucial to the total concept are not perceived to have an especially high priority. By setting ourselves a concrete goal we can focus the energies of the research community in a way that will ameliorate both of these problems. (2)

Wulf asked workshop participants to imagine the impact the collaboratory would have on the productivity of the nation’s scarce human resources. He argued that while it could be said that the collaboratory "would not enable anything new—anything not now possible, albeit more slowly or at the cost of moving people…" that, "the quantitative increase in ease of collaboration will have a profound qualitative effect" (2). That qualitative effect would:

• enhance the productivity of the individual researcher by providing access to information and instrumentation now available only at prohibitive cost in both time and money.
• increase the number of nimble minds and diverse perspectives far beyond those available at the researcher’s home institution.
• enable both inter- and intra-disciplinary research that simply isn’t being done now because the best people to do the research are not collocated and the scale or duration of the project does not justify a megacenter and the associated relocation of people.
• increase the pool of researchers available to work on the problem. The faculty at four-year and predominately minority institutions are an essentially untapped resource. The collaboratory will permit them to be full and effective partners in research projects, and increase the quality of instruction at those schools at the same time.
• speed the transition of new ideas into industry, into products—and increase the relevance of research to social/economic goals. By making academics and advanced developers in industry a part of the same collaboratory, the same ‘intellectual stew,’ we can achieve both effects simultaneously and naturally.

Wulf warned that we might never know quantitatively the impact of these combined effects because

we don’t know how to define ‘research productivity’ quantitatively, but also because we may not know what would have happened without the collaboratory. Moreover, not every researcher will wish to collaborate remotely, nor will any one single technological ‘fix’ cure the myriad problems faced by the country. (2)

Nevertheless, Wulf proposed, the effects will be profound. He explained that just one of the impacts outlined, the speed of technology transfer from research idea to product, had already been shortened from the usual 15-20 years to less than four years.

So, Wulf asked, "What needs to be done?"

For the most part we do not need to begin whole new areas of research. Rather, we need to coordinate and expand research already underway, to deploy the enabling infrastructure…to guide the establishment of national projects (such as the human genome and global-change databases) along paths that will permit them to interoperate, and, eventually, to set standards for both commercially-produced and one-of-a-kind instrumentation to be usable remotely. (2)

The first step, coordinating and expanding research underway, was to be undertaken at the Rockefeller University workshop. Wulf charged the workshop with setting the research agenda for the collaboratory, by addressing such questions as

• What are the central problems?
• What is the best mode for attacking each of them?
• How can research be coordinated?
• Who are good candidates for demonstration projects?
• How best can results be used as they emerge? (2)

Wulf concluded his White Paper with the recommendations that the NSF play a leadership roll implementing the agenda, but warned that the effort may become, in both size and scope, larger than any one agency can handle and will also require heavy involvement of the private sector (2).

Towards a National Collaboratory (Lederberg and Uncapher 1989)

The goal is to build no less than a distributed intelligence, fully and seamlessly networked, with fully supported computational assistance designed to increase the pace and quality of discourse, and a broadening of the awareness of discover: in a word, a collaboratory. (Lederberg and Uncapher 1989, 3)

The report of the Invitational Workshop Wulf convened at Rockerfeller University on March 13-15, 1989, Towards a National Collaboratory, endorsed the concept of the collaboratory enthusiastically. It recommends a "three-fold agenda that would repeat itself in cycles of design, implementation, and testing" (8). The first cycle concerned systems architecture and integration, and would examine ways to allow people and machines to use the collaboratory's components most effectively. The second cycle would evolve tools and technologies themselves, and the third would develop user-oriented testbeds to validate both technology and organization.

Lederberg and Uncapher's 18-page report has five sections with an Introduction. The sections are:

Science and Support for Collaboration in which the concept of team science is discussed within the framework of Lederberg’s "Epicycles of Scientific Discovery" model (Lederberg 1989).

Expected Impact in which the collaboratory’s potential amplification of future research is discussed with particular attention paid to the impacts on:

• Global Change research underway in the NSF, NASA, United States Geological Survey (USGS), and Department of Energy (DOE);
• The Human Genome project underway at NSF, The National Institutes of Health (NIH), and DOE; and
• Parallel Processing Research underway at NSF, DARPA, NASA, and DOE.

The Functional Collaboratory – which provides a functional description of, and assumptions about, the collaboratory.

Research Agenda in which details of the three-fold approach are provided.

Conclusions and Recommendations, which details the critical factors and issues to which attention must be paid.

This study examines the Functional Collaboratory, Research Agenda, and Conclusions and Recommendations sections of the Lederberg and Uncapher report in detail.

The Functional Collaboratory

What is a collaboratory? As we use the term here, it is the combination of technology, tools and infrastructure that allows scientists to work with remote facilities (co-laboratory) and each other (collaborat-ory) as if they were co-located and effectively interfaced. (Lederberg and Uncapher 1989, 6)

First, the authors warn, it is important to remember that scientists have always collaborated. Such collaborations include the processes of publication, co-authoring articles and attending conferences, sharing students, and working on teams at experimental facilities.

The norm of mutual scientific criticism is an intense form of intellectual collaboration, however antagonistic it may appear. Besides its cognitive utility, criticism is also indispensable to a rational system for the allocation of resources, tenured positions, research funds, and facilities. (6)

As science becomes more multidisciplinary and requires increased access to expensive and rare resources that are impossible to duplicate, the collaboratory will allow access to this remote instrumentation, "greatly expediting such scientific research: The collaboratory will provide seamless access to colleagues, instruments, data, information, and knowledge" (6).

Lederberg and Uncapher's (1989) report provides the intellectual foundation of the collaboratory. It relies on two distinct models: The first model is Lederberg’s "Epicycles of Scientific Discovery," which outlines the processes and phases that can and should be supported through the collaboratory. The second model is Mark Stefik’s "The Functional Collaboratory", a flowchart that shows how each of Lederberg’s research functions might be supported through the technical capabilities in the collaboratory.

For example, one critical function of scientific research, project organization and management, is examined. The functions required from a collaboratory to support coordination of action, joint design, and resource scheduling would include

email and directories, tools to support structured discussions, a digital library with appropriate search mechanism, user education and training tools, real-time computer supported multi-media teleconferencing, a remote experiment scheduler, and so on. These capabilities are in turn supported by a number of enabling technologies, such as networks, advanced human-machine interfaces, high resolution displays, and video compression techniques. Finally, underlying the collaboratory paradigm must be an understanding of how groups work together as well as offering seamless new technologies to enhance the availability of old knowledge, permit scientists new means of accelerating the pace of discovery and support the amplification of human intellectual capability. (7)

The Research Agenda

The recommendations for a research agenda outlined in Section Four of the Lederberg and Uncapher report are based on the following assumptions:

• The underlying communications network provides a high level of minimum capability (TI communication lines with 1.544 Mbps throughput and sufficient bandwidth [T3 or 45Mbps] to support anticipated load without delay; eventually achieving 1 Gbps throughput);
• Computing systems with relatively high capability (workstations with 10 Mips processor speed, 10 Mbytes memory, and 1000x1000 pixel color display connected to high performance computing with processor power of gigaflops and beyond;
• Infrastructure described above is made available to users/scientists. (7)

The research agenda focuses "on developing and demonstrating the technologies required to make the National Collaboratory a reality (8). The agenda recommends a three-fold approach:

1. Systems architecture and integration aimed at the system issues that allow people and machines to effectively use various technologies that are the components of the collaboratory;
2. Evolution of the underlying technologies and tools themselves; and
3. User-oriented testbeds coupled with better theories about the process of collaboration that are required to validate both the technical approaches and the overall system components allowing understanding of the requirements on such a system and the role that the various technological components play in the overall design.

The three-fold approach is envisioned as "an iterative cycle of design, implementation, and testing" (8).

1. Systems Architecture

The report admits that most of the research that needs to be done to actualize the collaboratory is in the area of system architecture and integration, but it echoes Wulf that the collaboratory is more than a set of tools.

It is a functional capability to improve scientific effectiveness by taking advantage of a broad set of resources, including but not limited to remote facilities and other scientists. Understanding the appropriate system architecture, where such architecture includes not only the underlying tools but also the people that are to use those tools, requires a dedicated effort. (8)

2. Available Technologies

Lederberg and Uncapher point out that there were already tools available, both from the research community and the commercial sector, on which to begin design research on the systems architecture outlined above. These tools (and their limits, and suggested areas for research) include:

Electronic Mail (which needs interoperability, graphics capability, privacy, and user support);

Electronic File Transfer, which was already well proven;

Remote Access and Control, including remote logon (in need of enhancements to assure access control and authentication for safety and security);

Shared Files (which allows the sharing of ASCII files themselves, but with limited ability to share information though such files, and in need of a higher level of functionality and standardization to include graphics and data research representations);

Database Access (ability to store and retrieve data from shared databases needs standardization);

Access Control and Authentication (security mechanism need to be adopted and refined);

Multimedia Mail (integration of graphics, sounds, spread sheets, scanned images into text requires powerful mechanism for interoperability);

Structured Interaction Support (migration of multi-media teleconferencing, structured conversations, and information sharing from proprietary platforms to open architecture);

Simulation of Instruments (prototyping instruments through computer simulations via totally compatible hardware and software with remote debugging capabilities; after which, the distribution of simulated instruments to large number of scientists for conduct of large, simulated distributed experiments).

The recommended research agenda would investigate the integration of these tools and the development of user interfaces in a context of user feedback (with the following objectives):

3. Integrating Technologies

Digital Instrumentation (dedicated to theories and technologies for development of real-time control and feedback from instruments, and real-time communication concerned with remote and multi-user instrumentation, experiments with communication and control technologies, development of interfaces to address delays and scheduling, and use of human and machine agents);

Multi-media Meetings (experiments on collaboration in meetings, distributed education; economic means for providing very high-bandwidth transmissions; social experiments to study the effects of technology on interpersonal argumentation at a distance);

Digital Mail (develop ability to send a verified and trustable electronic check, linking value-added services, better addressing mechanisms, comprehensive yellow-page services, technical extensions and social issues surrounding the use of electronic mail);

Scientific Reference Service (service to provide expertise and network to answer tough questions, a "Who Knows...?" service using literature access and intelligent agents, but drawing first on the human, then on artificial intelligence);

Digital Journals and Peer Review (services for logging documents, logging comments on documents, and offering digital document retrieval; support for collaborative writing, experimentation with different modes for commentary, editing, and document exchange; social experiments to determine salient effects on the perceived qualities of number of reviews, and effects on peer group size);

Digital Library (many variations including software, video, and other "unusual forms" in a distributed electronic database; collaboration and electronic publishing; integration of services over distributed libraries; social experiments with policies, citation mechanisms, pricing, collection and distribution of royalties);

The digital library was given added attention in the report’s section on integration of technologies. Of particular interest was development of techniques for discovery through digital search, including technologies for scanning and comparing strings of digital data; document storage; and content recognition capabilities through large-scale linguistic analysis or comparison. Social questions about the digital library include analysis of the ways that automated searching enable collaboration or discovery. "The digital library is likely to be the key to valuable old knowledge, and new knowledge so vital to the scientific process" (12).

4. Advanced Technologies

Developing the integrated capabilities outlined above, the report says, will require developing tools that did not yet exist. These underlying technologies are described as:

Hypermedia Conversion Support such as hypermedia databases to track design decisions, operational problems and corrections, and research approaches.

Intelligent Agents such as distributed processes that would act on behalf of users. The report visualized each entity in the collaboratory (scientists, instruments, databases, computer resources) having an intelligent agent between it and the network. These intelligent agents would act on behalf of the entity, negotiating with other agents, conducting searches, scheduling, etc.

Interoperable Data Description that would describe data from multiple disciplines using a common format to allow interoperable data analysis and manipulation.

Information Fusion that would allow integration of information from heterogeneous sources.

Smart Agents for the Design of Experiments to facilitate the use of multi-sensor experiments by multiple investigators.

Smart Data Gathering incorporating intelligence into the instrument that would allow "self-directed" data gathering (13).

Technology Utilization

How the emergent technology is used, and "the relationship between technologies and the way scientists do and will conduct their research" (14) is identified as a critical issue needing investigation in order to make the collaboratory a reality. Lederberg and Uncapher recommend that appropriate testbeds to understand the impact of technology be coupled with a research program into the underlying mechanisms of the collaboratory itself.

User Testbeds

Historically, many new technologies have been left dangling at the end of the research cycle waiting to be adopted by some user community or integrated into commercial projects (14). Lederberg and Uncapher suggest setting up user-oriented rapid-prototyping testbeds as partnerships between users and developers, and identified several critical attributes such testbeds must have:

• Represent a partnership between users who see the potential for the collaboratory improving their scientific research and technologists who are interested in working with such a community.
• Be of a size that is sufficient to explore the impact of the collaboratory on scientific research (which means a team of geographically dispersed users working with laboratory facilities) but small enough to be manageable as a rapid prototyping environment.
• Provisioned with adequate infrastructure (networking, workstations) so that the prototype does in fact represent future potential.

The report notes that it is important that those who are working on developing technologies be provisioned at any stage with the next stage of infrastructure so as to act as a leading edge for the technology.

Collaboration Mechanism

In order to develop tools for supporting scientific collaboration, the Lederberg and Uncapher report insists it is critical to understand the process of collaboration itself. Achieving this understanding is a multidisciplinary enterprise and will draw on a variety of existing disciplines. Some of the work will be observational (field studies, surveys, archival analysis), some will be experimental. The testbeds will be one environment for conducting this research. It is also crucial to synthesize previous work into new theories about how collaboration and coordination occur and how technology can help (15).

Examples of Research Questions

The Lederberg and Uncapher report provides a description of some of the questions that needed to be addressed by the National Collaboratory research agenda. Those questions include:

What are the basic processes involved in coordination?

What structures are possible for carrying out these processes?

How is collaboration among scientists different from other kinds of collaboration?

How might use of collaboration technologies affect incentive structures for the conduct of science?

Is the social science structure of science affected by intensive use of communication technology for remote collaboration and resource sharing?

To what degree is it possible to substitute capital (electronics) for scientific labor?

The Report's Conclusions and Recommendations

A number of factors and issues addressed in the conclusion of the Lederberg and Uncapher report bring firm focus to the critical elements necessary for developing the National Collaboratory:

• Importance of integration and user testbedding
• Careful selection of user testbed communities
• Community workshop to bring users and developers together
• Targeted integrated system that all works together
• Drawing on experience of other agencies/organizations

Conclusion

The foundation of the collaboratory is provided by three key documents. The first two of those documents were never published nor widely distributed, although they are widely cited, and are examined in detail in this chapter. The first document, William Wulf's White Paper (1988) puts forth the concept of the collaboratory, sets its philosophical foundation, and identifies the disciplinary and research focus needs of the National Collaboratory. Wulf projected that the collaboratory would be an interdisciplinary endeavor requiring a relatively equal contribution from a variety of disciplines.

The second key document, Lederberg and Uncapher's (1989) Towards a National Collaboratory, sets the intellectual foundation of the collaboratory, outlines the National Science Foundation's National Collaboratory research agenda, and identifies the topics and approaches of research needed. Lederberg and Uncapher reflect Wulf's suggestion that the collaboratory be constructed as an interdisciplinary environment from relatively equal contribution from multiple disciplines.