Technology and the Design of

Open-Ended Learning Environments

Michael Hannafin
University of Georgia


I labored for some time to develop a style useful for this forum. Should I take the classic scholarly approach, and attempt to guide your interpretation through carefully weaving relationships among related theory and research? I'm familiar with this style, but it's a form better suited to monologue than dialog. Besides, it's already out there and readily accessible. Perhaps establishing a list of frequently asked questions and responses would be helpful? For some it might be very helpful, but for others there would be no context to understand the meaning of the questions and/or responses. Besides, questions and responses (mine and others) should surface through discussion.

I had a recent opportunity to speak to a group of researchers about open learning systems. Most were polite and feigned listening, but the majority expressed far more interest in the how's and why's of my involvement than with the academic weaving of "scholarly" argumentation. In the same vein, I decided to share the experiences that influenced my work, not just to rehash the details of it.

Some preliminaries. What are open-ended learning environments (OELEs)? How do they differ from directed or "closed-loop" learning environments? At the risk of duplicating information available elsewhere, OELEs are systems designed to support the unique search and understanding needs of individuals. That is, they are not designed to teach particular content, to particular levels, for particular purposes; they are designed to support learners' attempts to understand for their own purposes. In effect, OELEs impose no particular pedagogical strategy or instructional sequence, but guide learners in invoking their own strategies and generating their own learning sequences.

What does technology afford the design of such learning environments? Technology provides the means through which wide-ranging resources can be varied multimedia resources into digital DNA, and provides a myriad of methods for re-organizing it into different "life forms." Technology provides powerful engines that enable flexible search strategies, and tools with which to connect, link, record, capture, manipulate. Technology embeds (or make available) various kinds of advice and/or support (e.g., scaffolding, strategic guidance, etc.) to aid learners in constructing understanding. OELE use is typically informed by problems, needs, dissonance, controversy, and other devices that induce perspectives to be strengthened, refined, or refuted.

Enough of the preliminaries. Now to the revelations.

Revelation 1:You are what you are trained to be, but that does not limit what you can be.

I was trained initially as an experimental psychologist, groomed in radical behaviorist traditions. I built mazes, starved rats to 75% of their normal body weight, shaped the responses of pigeons to approximate a "dance," and systematically altered reinforcement schedules. I later worked in public schools for eight years as a school psychologist. I administered test batteries to classify intellectual ability and diagnose specific learning disabilities, structured token economies to manage student behavior in classrooms, developed explicit reward contingencies to increase study time (the desired behavior, of course), and designed lots of exercises to help children discriminate between commonly confused alphabetical characters (e.g., "p" and "q" and "b" and "d"). I engineered externally controlled, directed learning environments, and was pretty content in doing so. The stuff worked. Life was good. I do not, as some do, view behaviorism as somehow inherently flawed; likewise, I do not categorically discount the value of my experimental research traditions. Quite the opposite, I see many very effective applications of both for teaching and learning. But they weren't enough for me.

Revelation 2:Kids say and do the strangest things.

I became a parent for the first time in the mid '70s. I was immediately struck by how my son, then other children, made sense of the world--not always "correct" interpretations, but sense nonetheless. I learned of the works of Jean Piaget, Sigmund Freud, and even B.F. Skinner, each of whom was acutely sensitive to the development of their children and/or grandchildren. Their research and theory was influenced heavily by these observations. I wish I had equivalent motivation, but I didn't. I was a wide-eyed first-time father who was mesmerized by the emergence of a young mind. I marveled, for example, at how early language evolved--the peculiar mix of modeled speech and inferred language rules (at a young age). After hearing words such as "like" become "liked" and "try" become "tried" in past tense, children derived tacit rules and spoke surprisingly "properly." They didn't know what past tense was, but they evolved understanding through experience--active and vicarious--that enabled them to generalize. They also applied their rules to form incorrect past tense, such as "eat" becoming "eated" and "fight" becoming "fighted."

I watched as my son began formulating his own ways to interpret and understand his world (spontaneously grouping objects by attributes, organizing multiple objects into smaller sets, etc.). Each child, it seemed, evolved his or her own ways to understand and deal with surroundings. Kids formed incredibly sophisticated and complex understandings--though not always accurate understanding--with hardly a hint of explicit external direction. Through interactions among things, ideas, and people, interpretations took new forms. Through experience, children evolved understanding; guidance during experience helped to support this evolution. Learning without direct instruction--what a concept! This, I concluded, was pretty amazing stuff; this was hard to ,gnore. Obvious to many, but not to me.

Revelation 3:Researching how people learn is not the same as researching if people can learn in particular ways.

During my rat-running days, I found that many kinds of learning and performance could be readily engineered. In schools where the reward contingencies were explicitly known to students, they learned even quicker. In fact, successful students were especially good at determining and demonstrating what and how they needed to do to pass my classes--that is, how to think (McCaslin & Good call this "compliant cognition"--wish I'd though of that term). Clearly, I could engineer performance using prescribed methods.

Yet, I also observed a great deal of learning and performing that was not engineered externally. I could make the pigeon to "dance," but that didn't make the pigeon a dancer. I could engineer thinking and learning, but I didn't understand how either occurred. I had little insight into how learning occurred spontaneously. As I watched both younger and older learners, I became interested in how individuals made sense of their world, and how learning environments might become better aligned with the individual's sense-making efforts. This has become known by various phrases (e.g., learner-centered design), but fundamentally the perspective was one of trying to create learning environments that capitalize more on how individuals reason than imparting particular kinds of reasoning.

Revelation 4:Frameworks are better for explaining than predicting.

My foray into open learning was not due to the compelling arguments about the flaws of behaviorism or the merits of cognitivism. It was not due to "trendy" research or theory and was not a response to the incredible capabilities of technologies. It was a personal reaction to what I saw around me. I observed different and very powerful learning processes in my own child, and I knew little about them. I didn't seek truth or to predict outcomes--I sought to explain. The need to better understand seduced me into philosophical and psychological territory about which I was largely unaware. Contemporary developmental (e.g., Piagetian), psychological (e.g., situated cognition), and philosophical (e.g., constructivism) thinking didn't shape my initial perspectives--they provided ways to interpret, explain, and understand them. For me, this has become far more important than being able to predict or engineer.

Revelation 5:It's better to understand real differences than to enforce artificial similarities.

Over the years, I had the opportunity to engage in critical (though not usually hostile or negative) discussions with several individuals who helped me to come to grips with the murkiness in my thinking (not resolve it, but acknowledge it). One particularly significant encounter was with a good friend and eventual colleague Walter Dick, a well-known Florida State University professor in the instructional design field. Roughly 10 years ago, I spent several minutes enthusiastically outlining my efforts--eventually described as open learning environments. After listening patiently, Walt calmly stated that my work wasn't instruction. It lacked, according to this eminent instructional theorist, the key features of well-designed instruction: clearly stated behavioral objectives, well-aligned practice, assessment of ongoing progress toward learning outcomes, and so forth. Initially, I was taken aback, then became defensive about the comment. My inclination was to blow the comment off as another attack from an "ISD dinosaur" who hadn't yet seen the light. Of course, Walt was neither a dinosaur nor incorrect. More importantly, he wasn't negative or critical--he just made a simple observation. Walt stated something that to him was obvious, but hadn't yet crystallized in my own mind: Instruction is neither the only, nor in many cases the best, method to support learning. This, I'm sure, wasn't Walt's intent, but it clarified something for me: I was interested in something different from instruction.

I felt liberated until I realized that I could not simply assume the same foundations and underlying assumptions I had before. If not instruction, then what? How did open learning environments relate to instruction? The differences between instruction and open-ended learning environments needed to be better understood and articulated. I had to re-assess the foundations of such endeavors and assumptions about the nature of learning, as well as to consider different perspectives on the structure of information and knowledge. It required a clearer delineation of how individuals might be induced to learn (when not explicitly directed), and about the various kinds of resources that might be relevant to such learning. Walter Dick's criticism was, for me, a defining moment. I was interested in something qualitatively different from instruction--not inherently superior or inferior--but definitely different.

Revelation 6:You can't ride a horse to the moon.

I am strongly influenced by Richard Clark's "mere vehicles" description of media, including computer-based technologies. However, I also find Bob Kozma's views about "learning with media" to be compelling. Both strike me as important, and not entirely incompatible, positions. I have tried to reconcile them because there is wisdom in each.

While at Penn State, Kyle Peck and I described a parable, the moral of which was that enhancing the performance of a horse can be very effective provided one's goals are attainable via a horse. Horses can be bred for speed and power, we can alter the conditions of a race track, and can improve performance via diet--all of which might make the horse faster. To the extent that the goals are consistent with the greater speed, we can reasonably expect that these enhancements will improve performance. However, not all goals are equally attainable with each mode of transportation. If we want to travel to the moon, we can't get there by riding a horse--any horse--no matter how well-engineered the diet or how selectively bred. We need to consider modes of transportation and methods that are better suited to the goal. Horses can't fly, but other means of transportation can. I resolved to try to clarify, in my own mind at least, what my goals were before I determined which "horse" to saddle up.

Revelation 7:There are several available design technologies.

The ability to apply design processes across various domains is key to the adoption and growth of such processes. This is where instructional systems design (ISD) has been especially successful. Reliable, identifiable, and portable design processes were made available for use across a variety of teaching-learning domains. Until recently, however, non-directive learning environments existed largely as case-examples. We could marvel (or criticize) the interesting approach manifested in each environment, but it was often unclear how they were developed, or more importantly if the processes could be applied on a broader scale.

There now exist design technologies for various open-learning approaches. For example, individuals involved in the design and development of problem-based learning environments (e.g., John Bransford and the Cognition and Technology Group at Vanderbilt, Roger Shank at the Institute for the Learning Sciences at Northwestern University) have detailed methods for anchored instruction and case-based reasoning. Others have prepared guidelines for discovery learning, situated learning, and a host of other more open learning approaches. In theory, then, each advocate of an alternate learning environment could generate sets of procedures to guide development.

Revelation 8:Revelation 7 complicates rather than simplifies things.

The elegance of most ISD approaches is that they are applicable to diverse domains, varied design modes (e.g., tutorials, drills, etc.), for learners of various ability, age, and background. We could argue the degree to which this has actually happened, but most would concede that conceptually and procedurally similar ISD approaches have been widely implemented.

A key problem with the trend in open learning, however, is that the raw number of "unique" methodologies has grown dramatically. Unlike ISD approaches (which, though numerous, embody similar generic processes which can be readily identified), there is as yet no unifying approach sufficiently robust to accommodate both the similarities and differences across approaches. Some might argue that this is as it should be--that a "one size fits all" design methodology should be abandoned in favor of multiple approaches that better reflect the assumptions of the specific approaches. Others suggest that, if such approaches are to find a significant place in training and education, we need better ways to distill the important aspects of diverse teaching-learning methods into a more powerful, versatile, and generalizable design model--akin, in impact, to generic ISD approaches. Personally, I waffle on the issue. I'm looking to distill into a unifying, generalizable design system, but I'm nowhere near accomplishing it. I believe it is essential to the broader adoption and application of open learning designs, but haven't yet "cracked the code" on how it should be done.

Revelation 9:Better to risk a grand failure than limit myself to small successes.

This is a value judgment on my part. Much of the truly important work in education, and in educational research, addresses relatively small problems in terms of scale and impact. My own research emphasized small, well-defined problems for many years. I was comfortable and had some success, but again found myself wanting. I kept looking for larger meaning. I recognized the perils of research designs that were not tightly executed, but I couldn't reconcile my own dissatisfaction. Things didn't exist in the neatly organized, controlled ways that I could contrive for research purposes; much of what I "proved" as a researcher fell apart in natural learning contexts (or, at least, worked very differently). There had to be more.

Revelation 10:Other people--lots of them--are pretty smart.

I consider this to be self-evident (I know many people who agree wholeheartedly). I find this to be especially true in areas where I don't know what I don't know. But it's also true in areas where I think I know something but know very little.

I've had the great fortune to work with (and to benefit from the work of) several very bright individuals--university faculty, graduate students, classroom teachers, trainers, researchers--who have shaped my beliefs. I marvel at how different perspectives on the same phenomena contribute to my own understanding (this was true even before the phrase "social cognition" was in vogue). I have tried to discover not only what others know, but about how they think and conceptualize their work. More than any single source, my thinking has been shaped by the varied, sometimes contradictory, opinions of very bright individuals.

Where does this leave us? The topic, as CNN talk-show host Larry King might say, is technology and open-ended learning environments. I invite you to share your thoughts--from a classical research and theoretical perspective if necessary, but from your experience if possible. Maybe we can learn more about how each other thinks as well as what each other says. Looking forward to your thinking--and saying.

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

The buttons that appear below will be found at the 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 #14). 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.