Gerry's Home Page Perspectives on CSCL Journals & Book Chapters Conferences Dissertations Tech Reports Presentations Proposals Interpretations Ideas Selected Writings


POW! Perspectives On the Web

1.    Identification

Title of proposal:

POW! Perspectives On the Web


Name of principal investigator:

Gerry Stahl


Name of university or institution:

University of Colorado



Department of Computer Science


Date of submission:

October 18, 1999


Technical committee:



2.    Research Objective

Summary: This project will develop effective systems to represent personal and group perspectives on the Web. Computational support for perspectives can significantly enhance the potential power and usability of Web-based collaboration systems for groups of learners or workers. This can help break down the barriers to widespread participation in computer-mediated collaboration for the development of ideas and allow computer-based knowledge-building communities to proliferate around the world.

Problem: Collaboration is highly recommended for education and is a fact of life in contemporary work. But collaboration is very difficult, particularly in situations that require the building of new knowledge by the collaborating group. Collaboration requires a complex interplay between individual and group tasks, that are hard to keep track of and to manage. Computer support of collaborative learning (CSCL) and of cooperative work (CSCW) promise important advantages, such as: independence from people having to be at the same place at the same time, persistence of ideas, tools for analysis and reflection, access by extended virtual communities. However, current systems like threaded discussion forums suffer from serious weaknesses in practice. Discussions tend to be short-lived, inspire low participation, result in divergence of ideas, and remain superficial.

Technology and Usage Issues: The proposed POW! Project addresses the barriers to technology adoption in an important area of information management: an area identified in this proposal as "collaborative knowledge-building environments." These systems promise to provide effective environments for education and knowledge work. They represent a new usage paradigm associated with the facilitation of effective online communities. The proposed system design approach is responsive to the social implications of the technology in typical use contexts.

Software environments to support collaborative knowledge building need to incorporate a well-designed set of computational capabilities and they need to be used within a carefully orchestrated set of social usage conditions. The POW! Project will design, implement, field and assess a technology of computational perspectives within a knowledge-building environment. This technology, when configured and used within the proper social conditions, will represent a set of personal and group perspectives of the user community. Users collaborating in the environment will be able to define their own perspectives on common topics, exchange those perspectives, and negotiate shared group perspectives.

A computational perspective is a restricted view into a larger information space. It is defined by an individual or group to capture content of interest to them and to modify that content in their won view. The POW! system will provide support for defining and elaborating alternative perspectives on ideas. Relationships can be established among the perspectives, so that they benefit from each other without disturbing each other. The software supports an arbitrarily complex graph of perspectives for groups, multiple layers of sub-groups, individuals and even alternative perspectives for a given individual. There is a mechanism of content inheritance whereby perspectives incorporate the content of higher level group perspectives, so that there is sharing of ideas and each perspective does not have to start from scratch. Ideas can be negotiated so that they migrate from personal opinions to shared understanding and group knowledge. The perspectives are supported within a Web-based environment for knowledge construction and collaboration that provides a range of useful functionality.

Research Hypothesis: Using computational personal and group perspectives to structure the display of shared information spaces can facilitate collaborative knowledge-building processes.

Verification: Analysis of the usage of software prototypes using Perspectives On the Web will evaluate the increased effectiveness of collaboration in knowledge-building processes. Studies of threaded discussion systems lacking computational perspectives (dePaula, 1998; Guzdial & Turns, forthcoming; Hewitt et al., 1998) have demonstrated their weaknesses with the following measures:

(a)   short threads: a tendency for discussions to die quickly,

(b)  low participation: a lack of motivation to collaborate,

(c)   few cross-references: little convergence of ideas,

(d)  superficial content: minimal depth of investigation.

The POW! Project will field test software prototypes in a variety of contexts during a three year period:

(1) college seminar,

(2) corporate training,

(3) academic research group,

(4) corporate work group.

The resulting discussion will be analyzed using a variety of methods, including:

(i)   structural analysis: quantitative statistics on thread lengths, participation levels, etc.,

(ii)  discourse analysis: qualitative analysis of collaborative exchanges,

(iii) content analysis: coding of contributions and implicit cross-references,

(iv) survey analysis: interviews and surveys of user experiences.

The results of the analyses i-iv under conditions 1-4 will be compared with previous studies without perspective support to verify (or falsify) hypothesized improvements of measures a-d.

Potential Impact: Organizing contributions to knowledge-building processes on the basis of personal and group perspectives – assuming it proves to be effective – could contribute to a breakthrough in computer support of collaboration. Collaboration provides a powerful – virtually necessary – approach to learning and working in the modern and future world. Breakthroughs in computer support of collaboration could result in widespread use of computer mediation at all levels of schooling, employment and community interaction. The impact will be particularly dramatic for effective distance education, interactive organizational memories and engaging virtual communities.

3.    Relationship To Other Research & Practice

Related Work: The importance of the concept of perspectives has long been recognized in the philosophical and theoretical literature (Boland & Tenkasi, 1995; Gadamer, 1960/1988; Heidegger, 1927/1996) .The PI’s dissertation work defined a perspectives mechanism within a stand-alone hypermedia system for designers (Stahl, 1993a) . In the following seven years, he developed this approach further, eventually implementing perspectives on the Web and the WebGuide  environment for collaborative learning (Stahl, 1995; 1999c; 2000; Stahl & Herrmann, 1999) . Other hypermedia systems using perspectives have stalled after generating initial enthusiasm (Boborow & Goldstein, 1980; McCall et al., 1990; Mittal et al., 1986; Nelson, 1981) . Educational knowledge-building environments represent an important research area currently, but none of them incorporate perspectives support for groups and individuals (Guzdial & Turns, forthcoming; Hendricksen, 1999; Pea, 1993; Scardamalia & Bereiter, 1996; Slotta & Linn, forthcoming) .

Uniqueness of This Work: Commercially available threaded discussion systems are now abundant; however, they suffer from serious weaknesses and need to be incorporated within richer computational contexts. Research prototypes of collaborative knowledge-building environments are part of a vigorous sub-field within CSCL and CSCW. Some of them use the term “perspectives” in the sense of alternative displays (e.g., textual versus graphical) of the same data. Others provide personal views of just one user’s own contributions. But none represent relationships among personal and group perspectives through an arbitrarily complex graph of perspectives (Stahl, 1999a) . The POW! Project takes the relation of personal and group perspectives as a key structure of collaboration, tries to provide a computationally powerful and intuitively useful representation of that structure, and assesses its effectiveness.

Use of Related Work: The POW! Project is firmly situated within the active research sub-field of Collaborative Knowledge-Building Environments (CoKnowBEs). The PI is currently leading a research seminar on this sub-field and organizing a related workshop at CSCL '99. In addition, he is the PI on a CILT seed grant from SRI International to facilitate collaboration within this research community. The goal is to foster interoperability among related prototypes and systems so that the POW! Project prototype can incorporate data and functionality from other developers and they can ultimately incorporate the functionality of perspectives.

While the POW! Project has a specific, unique focus around the technology of computational perspectives, it is an integral part of a much more general effort to advance a broad class of collaborative knowledge-building environments for learning and working. The technological and (even more so) the social research issues that must be solved for these systems to realize their tremendous potential are far too complex to be solved by an isolated research effort. Building upon the POW! Project, the PI and his colleagues are compiling a shared knowledge base related to CoKnowBEs, to promote data interoperability among CoKnowBEs, to eventually build interchangeable CoKnowBE components and to generally increase the level of collaboration among developers and researchers concerned about this class of systems.

Long-Term Research: The idea of computational perspectives was explored at Xerox PARC 20 years ago (Boborow & Goldstein, 1980) and the idea of collaborative knowledge-building environments has been pursued for 10 years (Scardamalia & Bereiter, 1996) . Yet, there are still important open research questions – both technical and social (Kling, 1999) . The Web has been wildly successful with applications that are structurally very simple, trivial to use and relatively easy to understand. Collaborative knowledge-building is inherently complex, requires subtle interpersonal skills and is hard to comprehend. While it seems inevitable that the world move rapidly toward ubiquitous networked collaboration, this will not happen as automatically as Web browsing, email correspondence or e-commerce. First, researchers will have to develop well-designed sets of functionality to support the structures and processes of collaboration and carefully orchestrated social procedures will have to be put into place. Significant progress in this direction can be expected in the next decade if efforts are made to address these issues systematically.

Research Strengths: The PI has been working on software incorporating perspectives for almost a decade. His WebGuide software has already been developed in an initial prototype and fielded in two use situations. He works within a research center that is well established and has been exploring the design of high-functionality software and computational support for learning and work for over 15 years. The PI is in working contact with researchers in the US, Canada, the UK and Germany engaged in related work. The Project will have access to appropriate test sites and will leverage supplementary funding for colleagues and equipment.

Complementary Work: The PI is currently involved in the following complementary projects:

¨     “Organizational Memory and Organizational Learning” in the Center for LifeLong Learning & Design explores computer-based group memories and the use of conferencing systems.

¨     “Articulate Learners” in the Institute of Cognitive Science explores the use of latent semantic indexing in educational technology, with an emphasis on rigorous testing and evaluation.

¨     “Symbiotic Computing” (proposed) in the Department of Computer Science explores pervasive networking of servers, desktop computers, laptops and very small distributed and embedded devices to support collaboration using shared information spaces.

¨     “Data Interoperability in CoKnowBEs” sponsored by CILT at SRI explores collaboration among researchers in North America, including the definition of XML standards for exchange of data among knowledge-building environments.

4.    Research Potential Impact

Expected Results: In addition to conference presentations, journal papers and a monograph reporting on the conceptual framework, software prototypes, field testing and effectiveness assessment, the following concrete products are expected:

¨     A PowServer Java middleware application that accesses the shared database, computes displays in specified perspectives and returns a tree of data to a client.

¨     A WebGuide Java client applet that provides a user interface in a Web browser to support collaboration.

¨     Analysis of field tests of the software in a variety of usage contexts.

¨     A set of XSL templates for displaying perspective data stored in a canonical format as an XML file.

Applications of Results: The PowServer and the XSL templates will be released as Open Source so that they can be used by anyone who develops a compatible client application. The WebGuide  client will be documented as an example Java client, but other clients based on HTML, Perl and CGI scripts could also be used. The PowServer API will be documented. The analyses of field tests will suggest how the perspectives software can be used effectively.

Benefits to the Computing Industry: The proliferation of more sophisticated, high-functionality applications on the Web will open up the Internet to a greater variety of uses and users. The potential of networked computation and collaboration will be further realized and this will drive the spread of ubiquitous computational devices.

Required Advances: Networking bandwidth (multi-megabit wireless access), computational power, memory capacity, device portability (size, weight, power consumption), multi-modal input and display interfaces (speech, large flat panel), seamless global roaming, acceptance of infrastructure of social practices. The technological advances are currently proceeding at a pace to make pervasive networked collaboration practical within the coming decade if the software design and social practice issues become well enough understood.

5.    Research Plan

The POW! Project is already underway. The basic perspectives technology is already designed (Stahl & Herrmann, 1999) . WebGuide, a preliminary prototype has already been developed and field tested in two educational contexts, a middle school environmental science class and a graduate cognitive science class (Stahl, 1999c) . An informal analysis has determined weaknesses that must be addressed in the next round of work (Stahl, 1999b) .

Experience to date indicates that there is no clear path to solving the complex social, educational and technical problems raised by this Project. A cyclical approach of thoughtful trial is needed. Each year, the proposed project will cycle through the phases of (a) theoretical reflection, (b) system prototyping, (c) practical application and (d) functional assessment. In Spring semester, the PI will conduct an interdisciplinary graduate seminar on the cognitive theory of knowledge-building environments. In the summer, the PI and colleagues will develop or revise prototype network-based software with functionality suggested by the theory and results of previous trials. In the Fall, the prototypes will be used in educational and work contexts. Assessments of the use of the prototypes will be conducted, published and incorporated into the next cycle.

Year 2000 Objectives: Assessment of the first WebGuide prototype revealed the following weaknesses:

(i)              Response time was too slow, especially on slow middle school computers at remote locations or for colleagues in Europe.

(ii)            Graph structure of perspectives and of discussion notes was not clear to users.

(iii)          Users were confused about which perspectives to use for discussion and for development of personal ideas.

(iv)          Despite functionality to permit and encourage convergence of ideas (topic structure, multiple linking, comparison perspectives), discussions diverged and died.

(v)            Data from different versions of WebGuide and from other similar software is incompatible and cannot easily be merged or compared.

The next version – WebGuide 2000 – therefore has the following year one objectives:

(i)              Separation of WebGuide into a PowServer running on a fast Web server to do intensive computations of perspective displays and a client applet to handle navigation within a perspective and to provide user interface services. Optimize Internet communication through the client/server API.

(ii)            Provide an optional graphical view of the perspective and note structure, with capabilities to add, delete, edit and link notes together. Increase the ability of the user interface to display large discussions.

(iii)          Revise discussion support and add support for private notes.

(iv)          Add negotiation support, including summarization, voting, debating and promotion to group perspectives.

(v)            Provide for import and export of data to a standard XML format and for display using XSL.


¨     A PowServer Java middleware application that accesses the shared database, computes displays in specified perspectives and returns a tree of data to a client. (Probably a revised version each year.)

¨     A WebGuide Java client Web application that provides a user interface in a browser to support collaboration. (Probably a revised version each year.)

¨     Annual analysis of field tests of the software in a variety of usage situations.

¨     A set of XSL templates for displaying perspective data stored in a canonical format as an XML file.

¨     Quarterly reports.

¨     Copies of works published as a result of this research.

Personnel: Only the PI will be supported from this grant. Related grants will support a graduate research assistant and an undergraduate research apprentice. A variety of colleagues, students and researchers at other institutions will collaborate with this Project indirectly.


¨     Release of a PowServer application running on a Web server.

¨     Release of a WebGuide client applet available on the Web.

¨     Field trial in a classroom, research project group, corporate training or business network.

¨     Analysis of field trial results.

Tradeoffs: A high-functionality Web-based application must always trade-off functionality versus ease of use and understandability. Scarce screen real-estate, response delays and limitations of different browsers and platforms further complicate software design. The proper balance requires experimentation and field trials to see what is most important and which problems can be tolerated.

Risk Management: The hardest problem will be to match customized versions of WebGuide and specially configured databases to field trial opportunities. Good opportunities come with fixed timetables, idiosyncratic priorities and a life of their own. It can take a lot of work to give the field trials a reasonable chance for success. The POW! Project will take advantage of a variety of kinds of trial opportunities, including some that the PI has considerable control over (e.g., his own seminars). The Project will develop a small number of high quality trial opportunities (one or two a year) rather than trying to take on more trials than it can adequately support and carefully assess.

Technology Exchange: The PowServer technology will be available to Intel engineers and they will have access to the databases of field trials. It is possible that in Year 3 a work group within Intel will want to try collaborating using the technology.

6.    Administrative Matters

Statement of Intel Access: Intel will be able to internally evaluate any prototypes developed during the research. The software prototypes will be maintained on the Web and Intel sponsors will be given passwords to review and evaluate the software and its use in field tests.

Researcher’s Name             Dr. Gerry Stahl, Research Professor


Researcher’s Organization: Center for LifeLong Learning & Design


Researcher’s Department  Department of Computer Science


Researcher’s Contact         Dr. Gerry Stahl, Department of Computer Science

                                                University of Colorado, Campus Box 430

                                                Boulder, CO 80309-0430

                                                (303) 492-3912 [voice], (303) 492-2844 [fax]



University Office of Contracts and Grants:

                                                Ralph Brown, Senior Contract & Grant Administrator

Office of Contracts and Grants, Campus Box 19

                                                University of Colorado

                                                Boulder, CO 80309-0019

                                                (303) 492-2692 [voice], (303) 492-6421 [fax]



Shipping Address                 Pat Libhart, Department of Computer Science

                                                University of Colorado, Campus Box 430

                                                Boulder, CO 80309-0430

                                                (303) 492-7514 [voice], (303) 492-2844 [fax]


7.    Personal Data

Website with Picture:

Resume                                            Education

                                    University of Colorado

                        1993    Ph.D. in Computer Science

                        1990    M.S. in Computer Science

                                    Northwestern University

                        1975    Ph.D. in Philosophy

                        1971    M.A. in Philosophy

University of Frankfurt

                        1973    Graduate study in critical social theory

University of Heidelberg

1968    Graduate study in continental philosophy

                                    Massachusetts Institute of Technology (MIT)

                        1967    B.S. in Humanities & Science (Math & Philosophy)


Research Professor

1999-present   Department of Computer Science and

Institute of Cognitive Science, Boulder, CO

Post Doctoral Research Fellow

1996-1999       Center for LifeLong Learning and Design, Boulder, CO


            1995-1996       Personalizable Software, Niwot, CO

                                    Director of Software R&D

            1993-1996       Owen Research Inc., Boulder, CO

                                    Graduate Research Assistant

            1990-1993       College of Environmental Design, Boulder, CO

                                    Intern Interface Developer

            1990-1991       US West Advanced Technology, Denver & Boulder, CO

                                    Computer Science Instructor & Teaching Assistant

            1989-1990       University of Colorado, Boulder, CO

Executive Director

1984-1989           Community Computerization Project, Philadelphia, PA

                                    Planning and Evaluation Specialist

1979-1984       Southwest Germantown Community Devel. Corp., Philadelphia, PA

Community Organizer & VISTA Supervisor

1978-1979       Philadelphia Council of Neighborhood Organizations, Philadelphia, PA

                                    Systems Programmer

            1974-1977       Temple University, Philadelphia, PA

            1970-1971       Northwestern University, Evanston, IL

            1969-1970       Temple University, Philadelphia, PA

                                    Applications Programmer

            Summer 1966  Brown Bovari Cie, Baden, Switzerland

            Summer 1965  University of Pennsylvania, Philadelphia, PA

Bibliography of Related Recent Publications:

(Stahl, 1993a; 1993b; 1995; 1999a; 1999b; 1999c; 2000; Stahl & Herrmann, 1999; Stahl et al., 1998; 1999; 1995a; 1995b)

8.    Proposal References:

Boborow, D. & Goldstein, I. (1980) An experimental description-based programming environment: Four reports, Technical Report No. CSL-81-3, Xerox Palo Alto Research Center, Palo Alto, CA.

Boland, R. J. & Tenkasi, R. V. (1995) Perspective making and perspective taking in communities of knowing, Organization Science, 6(4), pp. 350-372.

dePaula, R. (1998) Computer Support for Collaborative Learning: Understanding Practices and Technology Adoption, Masters Thesis, Telecommunications Department, University of Colorado, Boulder, CO.

Gadamer, H.-G. (1960/1988) Truth and Method, Crossroads, New York, NY.

Guzdial, M. & Turns, J. (forthcoming) Sustaining discussion through a computer-mediated anchored discussion forum, Journal of the Learning Sciences.

Heidegger, M. (1927/1996) Being and Time: A Translation of Sein und Zeit, (J. Stambaugh, Trans.), SUNY Press, Albany, NY.

Hendricksen, C. (1999) Asynchronous collaboration by Document Review, DocReview.

Hewitt, J., Scardamalia, M., & Webb, J. (1998) Situative design issues for interactive learning environments, at

Kling, R. (1999) What is social informatics and why does it matter?, D-Lib Magazine, 5(1).

McCall, R., Bennett, P., d'Oronzio, P., Ostwald, J., Shipman, F., & Wallace, N. (1990) PHIDIAS: Integrating CAD graphics into dynamic hypertext, European Conference on Hypertext (ECHT '90), pp. 152-165.

Mittal, S., Bobrow, D., & Kahn, K. (1986) Virtual copies at the boundary between classes and instances, Object-Oriented Programming Systems, Languages and Applications (OOPSLA '86), pp. 159-166.

Nelson, T. (1981) Literary Machines, Mindful Press, New York, NY.

Pea, R. (1993) The collaborative visualization project, Communications of the ACM, 36(5), pp. 60-63.

Scardamalia, M. & Bereiter, C. (1996) Computer support for knowledge-building communities. In T. Koschmann (Ed.) CSCL: Theory and Practice of an Emerging Paradigm, Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 249-268.

Slotta, J. & Linn, M. (forthcoming) The Knowledge Interaction Environment: Helping students use the Internet effectively. In M. J. Jacobson & R. Kozma (Eds.), Learning the Sciences of the 21st Century, Lawrence Erlbaum, Hilldale, NJ.

Stahl, G. (1993a) Interpretation in Design:  The Problem of Tacit and Explicit Understanding in Computer Support of Cooperative Design, Ph.D. Dissertation, Department of Computer Science, University of Colorado, Boulder, CO. Available at:

Stahl, G. (1993b) Supporting situated interpretation, Proceedings of the Cognitive Science Society (CogSci '93), Boulder, CO, pp. 965-970. Available at:

Stahl, G. (1995) Supporting Personalizable Learning, Technical Report No. CU-CS-788-95, Department of Computer Science, University of Colorado, Boulder, CO. Available at:

Stahl, G. (1999a) POW! Perspectives on the Web, Proceedings of the WebNet World Conference on the WWW and Internet (WebNet '99), Honolulu, Hawaii. Available at:

Stahl, G. (1999b) Reflections on WebGuide: Seven issues for the next generation of collaborative knowledge-building environments., Proceedings of Computer Supported Collaborative Learning (CSCL '99), Stanford, California. Available at:

Stahl, G. (1999c) WebGuide: Guiding collaborative learning on the Web with perspectives, Annual Conference of the American Educational Research Association (AERA '99), Montreal, Canada. Available at:

Stahl, G. (2000) Collaborative information environments to support knowledge construction by communities, AI & Society. Available at:

Stahl, G. & Herrmann, T. (1999) Intertwining perspectives and negotiation, International Conference on Supporting Group Work (Group '99), Phoenix, Arizona. Available at:

Stahl, G., Koschmann, T., & Ostwald, J. (1998) workshop: Shouldn't we really be studying practice?, International Conference on the Learning Sciences (ICLS '98), Atlanta, Georgia. Available at:

Stahl, G., Koschmann, T., & Scardamalia, M. (1999) workshop: Collaborating on the design and assessment of knowledge-building environments in the 2000's,  Proceedings of Computer Support for Collaborative Learning (CSCL '99), Stanford, California. Available at:

Stahl, G., Sumner, T., & Owen, R. (1995a) Share globally, adapt locally: Software to create and distribute student-centered curriculum, Computers and Education. Special Issue on Education and the Internet, 24(3), pp. 237-246. Available at:

Stahl, G., Sumner, T., & Repenning, A. (1995b) Internet repositories for collaborative learning: Supporting both students and teachers, Proceedings of Computer Support for Collaborative Learning (CSCL '95), Bloomington, Indiana, pp. 321-328. Available at:

9.    Budget

PI salary (6 months)                $ 41,200

PI fringe                                  $  9,700

Travel                                     $  3,000

Hardware & software             $  4,000

Computer support                    $  5,800

Total annual budget                 $ 63,700

Go to top of this page


Return to Gerry Stahl's Home Page

Send email to

This page last modified on January 05, 2004