Learning in Open-ended Environments:

Tools and Technologies for the Next Millennium

Michael Hannafin
University of Georgia

AUTHOR'S NOTE: This paper is a narrative version of the 1999 Peter Dean Lecture delivered at AECT's annual convention in Houston, TX. Portions of this paper are based on a chapter, Hannafin, M.J., Land, S., & Oliver, K. (1999). Open Learning Environments: Foundations, methods, and models. In C. Reigeluth (Ed), Instructional-design theories and models (Volume II). Mahwah, NJ: Erlbaum. This paper also draws from collaborative related work undertaken between the author several talented researchers (Hannafin, 1995; Hannafin, Hannafin, Land, & Oliver, 1997; Hannafin & Land, 1997; Hill & Hannafin, 1997; Land & Hannafin, 1997). Each helped to inspire the ideas presented in this paper, but bear no liability for its contents.

I have been inspired-even driven-of late by the phenomenal growth of interest in web-based teaching and learning. I am especially fascinated with and concerned about how flexible, resource-rich systems such as the web can be deployed to support user-centered learning. Given the vastness, ill-structuredness, and openness of such systems, how do individuals make sense of available resources for their own purposes? What methods can be provided to assist learners on fairly explicit tasks compared with needs that are spontaneously generated? To what extent can information systems be transformed in learning environments, and what mechanisms are needed to exploit both the capabilities of technology as well as those of individuals as they attempt to learn in open systems? These and related questions have been the focus of my recent work, and my collaborations with several others. This research is a work in progress; clearly, much has yet to be considered, evaluated, tested, and refined. What I report here is an overview of the state of our thinking about open-ended learning environments-warts and all. Of necessity, this paper will be light on details-there is simply way too much to present in this brief summary (I will provide additional examples, clarifications, and references as interest warrants). For now, though, consider the following foray into our world of open learning environments-their features, rationale, and potentials.

Open learning involves "...processes wherein the intents and purposes of the individual are uniquely established and pursued"; open learning environments (OLEs) "...support the individual's efforts to understand that which he or she determines to be important" (Hannafin, Hall, Land, & Hill, 1994, p. 48).

Openness refers to the learning goal(s), the means through which learning goals are pursued, or both learning goals and means. Goals can be externally specified, externally induced; or uniquely generated. In each case, though, the need to understand is established individually. The individual determines how to proceed based on his or her unique needs, perceptions, and experiences, distinguishes known from unknown, identifies resources available to support learning efforts, and formalizes and tests personal beliefs (Land & Hannafin, 1996). OLEs represent a way of thinking about the design and user needs of individuals who learn in open user-centered systems.

Components and Methods

OLE components include enabling contexts, resources, tools, and scaffolds. Enabling contexts orient the individual to a need or problem. They guide students in recognizing or generating problems to be addressed and framing their own learning needs. Externally imposed contexts specify the expected product of the learner's efforts as well as implicitly guide strategy selection and deployment. They are used widely where knowledge and/or skill accountability requirements are explicitly accepted as well as when concrete progress referents are deemed appropriate or necessary. They are often presented as explicit problem statements or organizing questions. The Great Solar System Rescue (1992), for example, places students in a dilemma in which a space vehicle has crashed on a remote planet. They are given clues, and challenged determine on which planet the crash occurred and the precise location of the crash site. The learner's tasks, for the most part, are explicitly delineated.

Externally induced contexts introduce a problem space but do not identify specific problems to be addressed. Rather, any number of problems or issues can be generated or studied at the discretion of the learner. The Jasper Woodbury Series (Cognition and Technology Group at Vanderbilt, 1992) features a dilemma confronted by the lead character, during which a situation is introduced where problems are apparent. The induced context introduces a circumstance that frames the problems or issues and solicits learner participation; the student interprets the context for meaning, generates sub-problems, and devises strategies based on individual interpretations of the enabling context. Students are provided perspective-setting or -altering contexts that help to activate relevant prior knowledge, experience, and skill related to the problem as well as help the learner to generate potential strategies to be deployed.

In generated contexts, the learner defines the enabling context based upon unique needs and circumstances. For example, a home gardener may wish to determine the cause and treatment of fungus growth in his or her vegetable patch. As with induced contexts, the generated context activates relevant knowledge, skill, and experience in order to frame problems and issues and to guide problem-solving strategies.


Resources, ranging from electronic (e.g., databases, computer tutorials, video), to print (e.g., textbooks, original source documents, journal articles), to human (e.g., experts, parents, teachers, peers), are source materials that support learning. The World-Wide Web, for example, enables access but the potential relevance of the available resources is often difficult for individuals to ascertain. A resource's utility is determined by its relevance to the enabling context and the degree to which it is accessible to the learner: the more relevant a resource is to an individual's learning goals, and the more accessible it is, the greater its utility.

OLEs make extensive use of available resources that provide an extraordinary reserve of source materials across a wide range of OLE applications. Resources can be either static or dynamic. Static resources do not change through use; information, such as photographic images of historical figures, is stable over time and is not subject to variation. Some media do not allow their contents to be altered such as the contents of videodisks, multimedia CD-ROMs, textbooks, and electronic encyclopedias. During learning, individual interpretations of resource meaning may vary and evolve, but the resource itself is unaltered.

Other resources change over time or through the introduction of new data, allowing the learner to repeatedly access the same resource but with different outcomes. For example, dynamic resources such as climatology databases created by the National Weather Service evolve continuously as daily weather data are entered.

Other media, such as The Human Body (Iiyoshi & Hannafin, 1996), contain both static and dynamic resources. The Human Body contains thousands of multimedia objects, including text, voice narratives, animations, digital movies, and graphic resources. The individual resources can be accessed or linked according to the unique needs of the learner. New information can be also linked to the resources in the form of personal notes, observations, and elaborations. The core resources remain intact, but dynamic functionality can be attained when users add to, revise, or otherwise customize contents according to perspectives and needs of individual learners.


Tools are the overt means through which individuals interact with resources and act upon their own thinking. Tool functions vary according to the OLE's enabling contexts as well as the intents of their users; the same technological tool can support different functions. Tools do not inherently enhance cognitive activity or skills in particular ways; rather, they provide a means through which thinking can be enhanced, augmented, and/or extended.

The first class of tools, processing tools, support the functions typically associated with information-processing models of human cognition (Iiyoshi & Hannafin, 1996). Seeking tools, for example, help the learner to locate and select relevant information. A variety of seeking tools exists, ranging from keyword searches, to topical indexes, to the semantic search engines available on the World-Wide Web. Collection tools are used to gather resources for individual purposes. They aid in the amassing of potentially important information which can be used to simplify subsequent access, support study in closer detail, or collect subsets of resources appropriate to individual learning needs.

Organization tools are used to represent relationships. Model-It, for example, assists learners in establishing, testing, and revising conceptual understanding. Model-It provides a graphical tool with which individuals can create and test qualitative models of scientific understanding (Jackson, Stratford, Krajcik, & Soloway, 1995a). General purpose organization tools such as Inspiration(tm) aid the learner in organizing and annotating concept maps depicting complex relationships. Integration tools link new with existing knowledge. The linking and constructing functions assists in both organizing ideas from a variety of perspectives and integrating them with personal knowledge.

Generation tools enable learners to create things. Several generation tools have been developed and studied. Iiyoshi and Hannafin (1996) described tools with which individuals create their own multimedia lessons. Rieber (1993) created a microworld where learners manipulate Newtonian physics concepts such as mass and velocity while attempting to dock a virtual spacecraft. RasMol (Raster Molecules) is an Internet-based learning environment used to create and display the structure of DNA, proteins, and small molecules (Sayle, 1996). Several RasMol shells can be downloaded and manipulated by learners. Molecules can be displayed as wireframe graphics, cylinder stick bonds, space-filling spheres, macromolecular ribbons, hydrogen bonds, or dot surfaces.

Communication tools support exchanges among learners, teachers, and experts. Synchronous communication tools support real-time interaction among participants. Asynchronous communication tools allow for extensive exchanging of ideas and or resources, but do not rely on the simultaneous availability of all participants. Examples of synchronous and asynchronous communication tools in practice are widespread. Blieske (1996) involved students in collaboration in the design of floor plans for a new home. Students shared asynchronously their design with other schools, collaborated on the merits of various approaches, then attempted to build each other's designs. Other projects involved students in different classrooms collaborating in writing different acts for a play (Schubert, 1997a), and writing stories for submission to on-line newsletters for publication (Schubert, 1997b).


Scaffolding supports learning efforts within an OLE. Mechanisms emphasize the methods through which scaffolding is provided, while functions emphasize the purposes served. Scaffolding complexity varies according to the locus of the problem(s) posed and the demands posed in the enabling context. When enabling contexts and problems are supplied, scaffolding can be closely linked to the domain under study; when enabling contexts are individually generated, scaffolding of a generic nature is generally provided. OLE scaffolding may or may not be faded as facility is attained. In externally imposed or induced contexts, for example, scaffolding may be faded since the nature of system needs and learner use can be reasonably established beforehand. For individual uses, where the nature of use and learner needs cannot be established in advance, scaffolding typically remains available but its usage becomes less frequent as the learner's facility increases.

Conceptual scaffolds are typically provided for externally-imposed or -induced enabling contexts, where it is possible to anticipate methods that are sensitive to demands. Conceptual scaffolds guide in what to consider. They help learners reason through complex problems as well as concepts where misconceptions are prevalent. At times, this is accomplished by identifying key conceptual knowledge related to a problem or creating structures that make conceptual organization readily apparent. Scaffolds can be provided through a variety of mechanisms, ranging from the graphical depiction of relationships among concepts, to outlines featuring ordinate-subordinate relationships, to information and hints provided by experts.

Metacognitive scaffolds support the underlying processes associated with individual learning management. They provide guidance in how to think during learning. Metacognitive scaffolding can be either domain-specific, such as where enabling contexts are externally induced, or more generic where the enabling context is not known in advance. It might remind learners to reflect on the goal(s) or prompt them to relate a given resource or tool manipulation outcome to the problem or need at hand. When a problem context is known, scaffolding can emphasize specific ways to think about the problem under. In contrast, scaffolding generic model-building represents a wide array of phenomena to be modeled with very different conditions. In such a case, the scaffolding focuses on the processes of creating models, including finding ways to link models with prior knowledge and experience, linking representational models to current understanding, and enabling learners to manipulate ideas through modeling tools (Jackson, Stratford, Krajcik, & Soloway, 1995b).

Procedural scaffolds emphasize how to utilize available resources and tools. They orient to system features and functions and otherwise aid navigation. Procedural scaffolding frequently clarifies how to return to a desired location, how to "flag" or "bookmark" locations or resources for subsequent review, or how to deploy given tools. The Human Body, for example, provides several resources and tools with distinct functions. Since the cognitive load associated with remembering all procedures for each tool and resource can be overwhelming, on-demand procedural demonstrations are available. Learners need not develop facility with all procedures until they have established, on an individual basis, the need for a given tool or resource.

Strategic scaffolds suggest alternative approaches during analysis, planning, strategy, and tactical decision-making. They help in identifying and selecting needed information, evaluating available resources, and relating new to existing knowledge and experience. The Great Solar System Rescue (1992), for example, offers a range of alternatives with varied degrees of direction. Probe questions offer explicit strategic clues for those needing a place to begin, while also helping to trigger a series of related strategies for those who are immersed in, but have not yet reconciled, a problem. Another strategic scaffold alerts the learner to potentially helpful tools and resources and provides guidance in their use.

Finally, strategic scaffolding may take the form of response-sensitive guidance at key decision points. For example, an individual might select a number of resources and "feel comfortable" with their understanding of concepts associated with gravity. Once an intention to exit the environment is indicated, she might be advised to test her understanding. She can make a prediction based on the perceived relationship between or among variables and test the prediction using manipulation tools.

Implications and Conclusions

The importance and utility of OLEs for both current educational practice and for emerging resource-based teaching and learning approaches cannot be overestimated. We need to optimize the utility of existing resources rather than continually recreate them. Billions of resources in diverse media have been produced during the past two millennia; this growth can only accelerate in the future. How can we not only make existing resources more available to support learning, but accommodate future developments in each?

We need greater utility from the resources we have and those on the horizon. We need systems that utilize resources flexibly, extensively, and efficiently. We need to accommodate diverse goals and needs involving identical (or similar) resources rather than redeveloping the same resources. The growth of both information and technology require that scaleable models be advanced, and designs that permit ready access, updating, and inclusion of growing bodies of resources. Yet, we must do more than simply afford better, dynamic access to rapidly emerging information systems. We need to stronger design technologies to optimizes rather than minimizes the reasoning capabilities of learners and support individual goals and needs. OLEs attempt to address these needs by inducing (or supporting) frames for study, making resources available, providing tools to support and encourage analysis and interpretation, and guiding learners in accomplishing their goals or addressing their needs.


Blieske, M. (1996). Collaborative design project for the Internet. [On-line]. Available: gopher://gopher.schoolnet.ca:419/00/grassroots/projects/final.reports/elementary/collaborative

Cognition and Technology Group at Vanderbilt. (1992). The Jasper experiment: An exploration of issues in learning and instructional design. Educational Technology Research & Development, 40(1), 65-80.

Great Solar System Rescue [Computer software]. (1992). Watertown, MA: Tom Snyder Productions.

Hannafin, M.J. (1995). Open-ended learning environments: Foundations, assumptions, and implications for automated design. In R. Tennyson (Ed.), Perspectives on Automating Instructional Design (pp. 101-129). New York: Springer-Verlag.

Hannafin, M.J., Hall, C., Land, S., & Hill, J. (1994). Learning in open environments: Assumptions, methods, and implications. Educational Technology, 34(8), 48-55.

Hannafin, M.J., Hannafin, K.M., Land, S., & Oliver, K. (1997). Grounded practice in the design of learning systems. Educational Technology Research and Development, 45(3), 101-117.

Hannafin, M.J., & Land, S. (1997). The foundations and assumptions of technology-enhanced, student-centered learning environments. Instructional Science, 25, 167-202.

Hannafin, M.J., Land, S., & Oliver, K. (1999). Open Learning Environments: Foundations, methods, and models. In C. Reigeluth (Ed), Instructional-design theories and models (Volume II). Mahwah, NJ: Erlbaum.

Hill, J., & Hannafin, M.J. (1997). Cognitive strategies and learning from the World-Wide Web. Educational Technology Research and Development, 45(4), 37-64.

Iiyoshi, T., & Hannafin, M. (1996, February). Cognitive tools for learning from hypermedia: Empowering learners. Paper presented at the annual meeting of the Association for Educational Communications and Technology National Convention, Indianapolis, IN.

Jackson, S., Stratford, S. J., Krajcik, J., & Soloway, E. (1995a). Model-It: A case study of learner-centered design software for supporting model building. Proceedings of the working conference on technology applications in the science classroom. National Center for Science Teaching and Learning, The Ohio State University, Columbus, OH.

Jackson, S., Stratford, S. J., Krajcik, J., & Soloway, E. (1995b). Making system dynamics modeling accessible to pre-college science students. Paper presented at the annual meeting of the American Educational Research Association, San Francisco, CA.

Land, S.M., & Hannafin, M.J. (1996). A conceptual framework for the development of theories-in-action with open-ended learning environments. Educational Technology Research & Development, 44(3), 37-53.

Land, S.M., & Hannafin, M.J. (1997). Patterns of understanding with open-ended learning environments: A qualitative study. Educational Technology Research & Development, 45(2), 47-73.

Rieber, L.P. (1993). A pragmatic view of instructional technology. In K. Tobin (ed.), The practice of constructivism in science education (pp. 193-212). Hillsdale, NJ: Erlbaum.

Sayle, R. (1996). RasMol. [Computer program]. Edinburgh, Scotland. University of Edinburgh.

Schubert, N. (1997a). Playwriting in the Round. [On-line]. Available: http://www1.minn.net:80/~schubert/Play.html

Schubert, N. (1997b). Signs of Spring. [On-line]. Available: http://www1.minn.net:80/~schubert/spring.html

Michael J. Hannafin, Director
Learning and Performance Support Laboratory
611 Aderhold Hall
University of Georgia
Athens, GA 30602

Phone: (706) 542-3157
Fax: (706) 542-4321
E-mail: hannafin@coe.uga.edu

The buttons that appear below will be found at the top and bottom of each page of the discussion. The first button will take you back to the previous page (in this case, to the beginning of paper #34). The middle button will take you to the ITForum home page. The last button takes you forward into the discussion as it progressed on-line.