Network Learning Environments and Hypertext:
Constructing Personal and Shared Knowledge Spaces

Michael J. Jacobson
James A. Levin

University of Illinois at Urbana-Champaign

Key words: networks, learning, knowledge, hypertext

Paper presented at the Tel-Ed '93 Conference, Dallas Texas

November 1993

Published in the Tel-Ed Proceedings as:
Jacobson, M. J., & Levin, J. A. (1993). Network learning environments and hypertext: constructing personal and shared knowledge spaces. In D. Foster & D. V. Jolly, Eds.), Proceedings of Tel-Ed '93 (pp. 190-197). Dallas, Texas: International Society for Technology in Education.


Educational uses of networks are rapidly expanding as the problems of "ease-of-use" and "access" are gradually being solved. However, even as these problems are being solved, the solutions create second-order problems, such as students and teachers becoming overwhelmed with too much network generated information. In this paper, we present conceptual frameworks that capture some of the unique properties of network learning environments, and which can be used to provide systematic guidance to the design of network learning activities and software tools to enable these activities. We illustrate these frameworks by showing how they have helped us design two different tools for educational uses of networks, a Message Assistant and a Project-Based Learning Server.


Considerable progress is being made in the development of a computer network infrastructure that can serve education. There is also research that suggests ways that distributed electronic networks may be employed in instructionally beneficial ways (Bruce & Peyton, 1992; Hunter, 1992; Levin, Riel, Miyake, & Cohen, 1987; Newman, Goldman, Brienne, Jackson, & Magzamen, 1989; Riel, 1990). However, as researchers and educators begin to consider the more widespread utilization of network-mediated resources, the very richness of these resources imposes special logistical and cognitive demands on the user that may unfortunately mitigate the learning promise of educational network use (Riel & Levin, 1990; Ruopp, Gal, Drayton, & Pfister, 1993). A critical area for research involving network learning environments (NLEs) is this interface between the educational user and network learning environment. Of course, software tools must be easy and relatively transparent to use by a wide range of students and teachers, and great strides have been made in software that is available to use and access network-mediated resources such as electronic mail, electronic bulletin board systems, and information servers. But the increasing power and ease-of-use of these network software programs is a two-edged sword. On one hand, students and teachers can now easily communicate electronically and access a wide range of information resources. On the other hand, with a huge mega-network like the Internet (10 to 20 million users and over a million servers), a teacher may receive hundreds of messages in a day or have trouble deciding on which of the million Internet servers to look for a particular piece of information.

A new class of NLE tools is needed which preserves easy utilization of network resources, yet helps manage and filter unnecessary complexity that interferes with network-based learning activities. These tools need to be designed based on an understanding of the special characteristics of electronic networks to help support the attainment of substantive learning goals such as problem solving and independent thinking. We have proposed a distributed network learning framework that can describe characteristics of information flow on small and large networks. These characteristics, in combination with the speed and size of the network infrastructure being developed, constitute special conditions for a learning environment that are quite different from the traditional classroom and home learning environments. A systematic conceptual framework is needed that can guide the design of educational network tools.

In this paper, we first review our earlier work on a perspective for understanding distinctive characteristics of electronic networks that we call the Distributed Network Learning Framework (Levin & Jacobson, 1992). Next, we discuss a hypertextual "knowledge spaces" conceptual model which we feel has applicability for the design and use of NLE software tools (Jacobson & Levin, in press; Levin & Jacobson, 1993). This paper concludes with a discussion of research we are conducting that involves developing and studying the use of software tools based on the Distributed Network Learning Framework and the Knowledge Spaces conceptual model.

The Distributed Network Learning Framework

In our research on educational uses of networks, we have observed the flow of knowledge across the network, assisted by the active involvement of network mediators (Levin et al., 1987). We have begun to characterize these observations as a Distributed Network Learning Framework (Levin & Jacobson, 1992). In this framework, information and knowledge flow from where they currently are to where they are needed. The focus of our work with this framework has been to look at ways to structure networks, especially the mediators, so that learning can take place most effectively.

Why would anyone learn something? From the perspective of this framework, one reason people learn something is because it is easier than regenerating the information the next time they need it. As an example, people do not memorize every telephone number they call. If they are calling Joe's Pizza, they look it up in the phone book. The next time they call for a pizza, they might also look it up. If they get pizza a lot, they may write the number on a piece of paper and stick it next to the telephone. After a while, if they are real pizza fans, they will know the number to Joe's Pizza by heart. The phone number information thus moves from the phone book, to the piece of paper by the phone, and finally to the person's memory.

This phenomena is even clearer with networks. The need to memorize (i.e., locally store) information is reduced if the ability to access it becomes easier. As networks and information technologies become more wide-spread and integrated into the world of work, the goals of education are shifting from knowing a fixed body of knowledge to knowing how to think independently and how to find information when needed. Furthermore, we live in a period of rapid change, in which knowledge that was true yesterday may not be true today (e.g., Joe's Pizza may move and get a new telephone number). Thus, "facts" are taking on an increasingly fluid and changing quality. This simple example illustrates the distinctive elements of educational networks related to the flow of information that are described by the Distributed Network Learning Framework. [1] These features are shown in Table 1.

Distributed Network Learning Framework    Example Comments

* Information flows through networks Pizza lover (i.e., the mediator) decides based on decisions made by mediators initially to look up the number, then to (students/teachers or a computer agent) write it down on a note, then to at each node. remember it.

* Information and knowledge flow toward The phone number moves from the phone where the learners need it. book, to the note, to the pizza lover's memory.

* Information appearing at a network Pizza lover makes different decisions node is stored locally if the mediators about locally storing information, first expect the information to be of value. looking it up, then writing it down, and finally learning it. The phone number would not have been stored if the pizza turned out to be bad and thus that piece of information judged to be of little value.

* Over time, mediators at nodes would The pizza lover gradually optimized her optimize the organization of locally ability to access the pizza phone stored information. number, as it was initially slow access in the phone book, then faster access with the written note, and fastest when committed to memory.

* Network-based information is not a Joe's Pizza may get a new phone number. static, fixed "thing" but rather is dynamic, fluid, and changing.

Table 1. Features of the Distributed Network Learning Framework.

Knowledge Spaces, Hypertext, and Network Learning Environments

But how are we to convey these distinctive--but unfortunately somewhat complex and abstract--features of network learning environments to students and teachers? Given research on the value of an appropriate conceptual model to assist users in operating a complex device or computer program (Norman, 1988; Norman & Draper, 1986), we have been developing software tools for NLEs that employ a Knowledge Spaces conceptual model. The Knowledge Spaces conceptual model possesses a number of analogical components that can function at several levels relevant to NLEs. At the core of this conceptual model is a spatial analogy that has useful target aspects for describing important characteristics of electronic networks (e.g., "objects can be physically located in different places separated by space" maps to NLEs where "computers and people are physically located in different places and separated by space"). The Knowledge Spaces conceptual model also suggests other familiar systems that can serve as a coordinated set of analogies for describing NLEs. For example, "highway networks that connect physically separated places and people for transportation purposes" are analogous to "electronic networks that connect distributed computers for information transmission purposes."[2]

The Knowledge Spaces conceptual model can also be used to articulate abstract epistemic notions about "knowledge structure" and constructivist approaches to learning that are relevant to instructional uses of NLEs. For example, Wittgenstein (1953) employed the analogy of knowledge-as-a-landscape in the preface to his Philosophical Investigations, while a criss-crossing-a-knowledge-landscape (inspired by Wittgenstein) has been used as an analogy for research into a constructivist conception of the nature of learning (Jacobson & Spiro, 1993; Spiro, Vispoel, Schmitz, Samarapungavan, & Boerger, 1987). A NLE may thus be "viewed" as a knowledge space that can be explored and criss-crossed for different purposes and from different conceptual perspectives, with different learning possibilities afforded by each traversal.

Another aspect of "knowledge spaces" is that they exist along a continuum from personal to shared (Jacobson & Levin, in press; Levin & Jacobson, 1993). Personal knowledge spaces are constructed for one's individual learning and knowledge utilization purposes (e.g., personal electronic mail messages or a personal "knowledge-base" intended only for a single person's reference or future use). In contrast, shared knowledge spaces are created for information and knowledge dissemination involving larger audiences (e.g., electronic mailing lists, bulletin board news groups, distributed information servers).

Hypertextual tools are ideally suited for constructing the conceptual interconnectedness that is central to our notion of personal and shared knowledge spaces (Jacobson & Levin, in press; Levin & Jacobson, 1993).[3] The core notions of hypertext--flexible nonlinear and random access linkages between nodes of information--are at once simple yet very generalizeable to issues that are relevant to NLEs, such as flexible, nonlinear storage and access to information (Conklin, 1987). Also, there is beginning to emerge research that is suggestive of ways personal hypertext and hypermedia learning environments can help students see important conceptual interrelationships and to better learn complex knowledge (Beeman, Anderson, Bader, Larkin, McClard, McQuillan, & Shields, 1988; Jacobson & Spiro, 1993; Jonassen & Wang, 1993; Lehrer, 1991). Research into the characteristics of hypertextual learning should help inform our understanding of how to better design and use NLEs to enhance student learning. As we describe below, an area of research interest we are pursuing relates to applying design parameters for personal hypertext learning environments to distributed and shared network learning environments.

Theoretical Frameworks and Research Projects

The Message Assistant. Some of the implications of the Distributed Network Learning Framework and the Knowledge Space conceptual model will now be illustrated by two of our ongoing research projects. The first involves The Message Assistant, an electronic mail system intended to function as a software mediator helping to control the flow of network information in the form of electronic mail. In addition to the standard electronic mail features (e.g., creating, sending, and receiving messages), this program also allows the user to create a rule-base and hypertextual links that filter and organize the information in electronic mail messages. The user-defined rules and hypertext linking features can be combined to automatically construct a hypertext that interconnects information in different messages. (For a fuller technical discussion of the user-defined rules and other features of the program, see Levin & Jacobson, 1992.)

The user-defined rules check text strings in different message fields and will activate a rule if there is a match. The message actions include raising or lowering the message priority, creating a automatic reply or forward message, and automatically creating fixed and variable hypertext links between messages. The Message Assistant can create either fixed hypertext links or variable hypertext links. Fixed hypertext links connect two messages and can be created manually by the user or automatically by the program during certain operations (e.g., replying or forwarding a message creates a link between the original message and the new message). In contrast, variable hypertext links exist only under certain conditions and are used to link messages based on multiple topics, concepts, or themes (Jacobson & Spiro, 1993; Levin & Jacobson, 1992). We refer to these variable topical or conceptual links between different messages as message views, with each message view being a self-contained collection of hypertext links between messages.[4] As different messages may be relevant to several different topics or conceptual areas, the Message Assistant allows the user to create multiple views.[5]

As students will typically collect a large number of messages received while they are involved with network-based learning activities, the user-defined rules and the hypertext features of the Message Assistant are intended to assist the students as they construct a personal knowledge space from this information. Note that creating such a personal knowledge space is much more difficult to accomplish with a more traditional electronic mail program that merely lists messages that have been received or sent, or even impossible given the "read-and-delete strategy" that is common among many users.

Project-based learning server. Hypertextual tools are ideally suited for structuring multiple organizing frameworks for any given set of information. This type of functionality is especially important when a wide range of people need to flexibly access and use information, as is the case with a network information server. However, merely providing flexible access to information using hypertextual tools is not sufficient to ensure substantive learning will occur.

As we noted above, there is a body of research accumulating that is starting to identify the theoretical and design characteristics of effective hypertextual learning environments, such as: active, nonlinear student exploration of hypertext information nodes, student creation of new links in an existing hypertext, explicit depiction of important interrelationships between surface and structural knowledge components across multiple case examples, and student authorship of hypertext materials (Beeman et al., 1988; Jacobson & Spiro, 1993; Jonassen & Wang, 1993; Lehrer, 1991). In another area of research, we are investigating the application of important hypertext and hypermedia research findings such as these to the development of a network hypertextual project-based learning server. The server, currently under development, will contain a large knowledge-base of science education projects, science curriculum units, and science questions which will be coded in terms of dimensions such as underlying science principles and themes, age and grade levels, instructional tips, needed materials/supplies, duration, and so forth. The project-based learning server will also be designed to reflect current cognitive learning research into naive student models of different science topics, to suggest instructional techniques to promote conceptual change in students, and to demonstrate important interrelationships between scientific principles and concepts across different content areas. Also, the knowledge and activities in the server will be selected for their relevance to real world issues and problems.

Overall, the project-based learning server is intended to be a shared knowledge space that is structured in ways to help support enhanced student learning of complex knowledge as students are involved with a wide range of authentic project-oriented learning activities. We will also investigate how hypertextual tools for constructing personal knowledge spaces, such as the Message Assistant, can be used by students in conjunction with the shared knowledge space of the project-based learning server.


In this paper, we have presented some conceptual frameworks for systematically developing network-based learning environments that can help students and teachers as they jointly construct personal and shared knowledge spaces. We have described a software tool for constructing a personal knowledge space, Message Assistant, and we have sketched out our initial efforts at applying similar concepts and tools to the design of distributed project-based learning servers that can be used to construct shared knowledge spaces.

For networks to have a substantial positive impact on education, "ease-of-use" and "universal access" are not enough: students and teachers need conceptual frameworks to help organize their activity, they need tools that are consistent with such frameworks, and they need mediators to enable the activity. This paper describes some initial theoretical and research steps toward these goals.


This material is based upon work supported by the United States National Science Foundation under Grant No. RED-9253423. The United States government has certain rights in this material. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. This work was also supported by a grant of equipment from Apple Computer, Inc.


Beeman, W. O., Anderson, K. T., Bader, G., Larkin, J., McClard, A. P., McQuillan, P. J., & Shields, M. (1988). Intermedia: A case study of innovation in higher education (Final report to the Annenberg/CPB Project). Providence, RI: Brown University, Office of Program Analysis, Institute for Research in Information and Scholarship.

Bruce, B. C., & Peyton, J. K. (1992). A situated evaluation of computer networking to teach writing (Technical Report No. 565). Urbana: University of Illinois, Center for the Study of Reading.

Conklin, J. (1987). Hypertext: An introduction and survey. IEEE Computer, 20(9), 17-41.

Feltovich, P. J., Spiro, R. J., & Coulson, R. L. (1989). The nature of conceptual understanding in biomedicine: The deep structure of complex ideas and the development of misconceptions. In D. Evans & V. Patel (Eds.), The cognitive sciences in medicine (pp. 113-172). Cambridge, MA: MIT Press (Bradford Books).

Hunter, B. (1992). Linking for learning: Computer-and-communications network support for nationwide innovation in education. Journal of Science Education and Technology, 1(1), 23-34.

Jacobson, M. J., & Levin, J. A. (in press). Hypertext and network-based learning environments: Technology for the construction of personal and shared knowledge spaces. Paper to be presented at the International Conference on Computers in Education: Applications of Intelligent Computer Technologies, Taipai, Taiwan.

Jacobson, M. J., & Spiro, R. J. (1993). Hypertext learning environments, cognitive flexibility, and the transfer of complex knowledge: An empirical investigation. Paper submitted for publication.

Jonassen, D. H., & Wang, S. (1993). Acquiring structural knowledge from semantically structured hypertext. Journal of Computer-Based Instruction, 20(1), 1-8.

Krol, Ed. (1992). The Whole Internet User's Guide & Catalog. Sebastopol, CA: O'Reilly & Associates.

Lehrer, R. (1991). Authors of knowledge: Patterns of hypermedia design. Paper presented at the annual meeting of the American Educational Research Association, San Francisco, CA.

Levin, J. A., & Jacobson, M. J. (1992). Towards a distributed network learning framework: Theory and technology to support educational electronic learning environments. Proceedings of the Fourteenth Annual Conference of the Cognitive Science Society (pp. 927-932). Hillsdale, NJ: Erlbaum.

Levin, J. A., & Jacobson, M. J. (1993). Educational electronic networks and hypertext: Constructing personal and shared knowledge spaces. Paper presented at the annual meeting of the American Educational Research Association, Atlanta, Georgia.

Levin, J. A., Riel, M. M., Miyake, N., & Cohen, M. (1987). Education on the electronic frontier: Teleapprenticeships in globally distributed educational contexts. Contemporary Educational Psychology, 12, 254-260.

Newman, D., Goldman, S. V., Brienne, Jackson, I., & Magzamen, S. (1989). Peer collaboration in computer-mediated science investigations. Journal of Educational Computing Research, 5(2), 151-166.

Norman, D. A. (1988). The design of everyday things. Reading, MA: Addison-Wesley.

Norman, D. A., & Draper, S. W. (Eds.). (1986). User centered system design: New perspectives on human-computer interaction. Hillsdale, NJ: Lawrence Erlbaum.

Riel, M. (1990). A functional analysis of educational telecomputing: A case study of learning circles. Paper presented at the annual meeting of the American Educational Research Association, Boston.

Riel, M. M., & Levin, J. A. (1990). Building electronic communities: Success and failure in computer networking. Instructional Science, 19, 145-169.

Ruopp, R., Gal, S., Drayton, B., & Pfister, M. (1993). LabNet: Toward a community of practice. Hillsdale, NJ: Lawrence Erlbaum.

Spiro, R. J., Feltovich, P. J., Coulson, R. L., & Anderson, D. K. (1989). Multiple analogies for complex concepts: Antidotes for analogy-induced misconception in advanced knowledge acquisition. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 498-531). New York: Cambridge University Press.

Spiro, R. J., Vispoel, W. L., Schmitz, J. G., Samarapungavan, A., & Boerger, A. E. (1987). Knowledge acquisition for application: Cognitive flexibility and transfer in complex content domains. In B. K. Britton & S. M. Glynn (Eds.), Executive control processes in reading (pp. 177-199). Hillsdale, NJ: Lawrence Erlbaum.

Wittgenstein, L. (1953). Philosophical investigations. New York: Macmillan.

[1]See Levin and Jacobson (1992) for a more complete discussion and rationale for various components of the framework. There are actually three main components of the framework: (1) network mediators and the flow of information and knowledge, (2) networks and cognitive theories of learning, and (3) the human-network interface. Given the space limitation here, we only discuss the first component.

[2]The use of multiple, coordinated analogies for network learning environments is important as it has been found that a single analogy can misrepresent important facets of a complex topic (Feltovich, Spiro, & Coulson, 1989; Spiro, Feltovich, Coulson, & Anderson, 1989).

[3]It is interesting to note that much of the most innovative network software development is using a hypertextual orientation either implicitly (e.g., TurboGopher's Bookmarks which are simple hypertext links to Gopher nodes) or explicitly (e.g., World Wide Web which creates very elaborate documents with hypertextual links to distributed network text and multimedia nodes (Krol, 1992)).

[4]We use the term variable hypertext links technically to describe message views as the hypertext links between messages that vary depending on the view the user has selected.

[5]We use the term views to be consistent with--indeed, to evoke--the knowledge spaces conceptual model. Ideally, a teacher would use the knowledge spaces analogy with her students as part of general discussions and activities involving NLEs, and the term "views" in the Message Assistant program would simply be a subtle reaffirmation of the general conceptual model.