Table of contents

Introduction.............................................................................................................................................. 1

Hypertext Features 20

Hypertext: History and Meaning 50

Ambron, Sue-Ann and Hooper, Kristina Multimedia. Redmond, WA: Microsoft Press, 1990, In Preparation.

Jeff Conklin Hypertext: An introduction and survey,
IEEE Computer Magazine Vol. 20 No. 9 1987, pp. 17-41.

David Jonassen and Nancy Roebuck HYPERTEXT/HYPERMEDIA Educational Technology Publications: Englewood Cliffs, NJ 07632, 1989.

Hypertext '87 Proceedings. November 13-15. Chapel Hill, North Carolina. Association for Computing Machinery: New York, New York, 1989.

Hypertext '89 Proceedings. November 5-8. Pittsburgh, Pennsylvania. Association for Computing Machinery: New York, New York, 1989.

Nielsen, J. Hypertext and Hypermedia, Academic Press: New York, New York, 1990 (In preparation).

Nelson, T. Literary Machines. San Antonio, Texas,1981.

Shneiderman, B. & Kearsley, G. Hypertext Hands-On! New York: Addison-Wesley, 1989.

Introduction 4

Hypertext, hypermedia, artificial intelligence and other advanced technologies are quickly captivating many with their inviting and implicit promise of a seemingly easy and intuitive environment for creating effective instruction, and even better, delivering instructional materials that are enjoyable and even exciting. These claims and promises are substantially accurate, and this book will not only try to promote these claims, but better than that, try to help fulfill their promise. It will go beyond the hype, help dispel the hype, and venture to create an easily approachable access path to these technologies.
Hypertext, hypermedia, and artificial intelligence are all empowering technologies. Instead of controlling a computer's actions through the unwieldy, but powerful, syntax of a computer language, they offer the promise and oftentimes the reality of creating expressive computer interaction in ways that are much more like building and constructing with paper and everyday materials. Instead of writing complex algorithms in an obtuse computer language, users of these new technologies can design the environment by building it while interacting with it and playing with it. "Playing" is the operative word, because the enriched power of these tools makes the act of creation much more spontaneous, and a natural outgrowth of expertise in non-computer areas, instead of a planned consequence of purely computational skills.
More than ever before, these technologies encourage and expect a user to employ her abilities and knowledge acquired in a professional or technical domain, and minimize the overhead of skills needed just to deal with a computer.

However, they still have a cost: they may force teachers, trainers, media experts, and even students to learn many new irrelevant things; but there is a grand payoff to offset the effort needed to overcome these hurdles. The payoff is twofold: an expanded ability to create and invent marvelous new things; and the satisfaction of gaining control of a powerful medium for expressing everyone's expertise in novel and fundamentally useful ways.
These technologies are all creative forces. They are cooperative agencies, enabling people with useful knowledge to create products that inform, entertain, and elevate others. They are more than media for communication, because they raise communication (connections, contact, conversation, touch) to new levels of sharing. Not only are ideas shared in abstract lexical representations, these technologies enable communication in a continuum of dynamic forms that are constantly being explored and enlarged. Nowhere is this more obvious than in the broad and fluid distribution of shareware over local networks of electronic bulletin boards and more conventional means of advertising and mailing. The concept of shareware had its beginnings in electronic games and simple utilities, but it may well become the core of hypermedia conversations and exchanges. Inventiveness is so explosive in this new area, that even the terms to describe its activities seem to pop up overnight and be in everyone's conversation: software, shareware, coreware, vaporware, everywhere.

The New Gutenberg and Columbus

This book is very much about exploration in many realms. It is about an adventure not into geographical space but into conceptual and metaphoric space. In fact, some people believe that the space already exists to be mapped out and discovered, while others are concerned with creating it anew. My gut feeling is that both things are occurring. We are discovering a hyperspace of concepts and conceptual structures that already exist in some abstract form, for much of what we know as knowledge, but for some knowledge, we are beginning to understand that the human mind is a very primitive analyzer and organizer of information, so that we have to create new and untrusted knowledge structures. Delving into the structure of knowledge, the jungle of concepts, skills, and instincts, to find the right simulations, metaphors and knowledge spaces so that this brave new world can be explored in safety; is as exciting and awe inspiring an adventure as any that have occupied mankind in its long evolution.

Some of the equipment and tools for this exploration even make the explorer adopt the helmets and goggles that give him the appearance of byegone eras, or futuristic fantasies. Body suits, eyephone goggles, datasuits, datagloves, and other rarefied paraphernalia let the conceptual adventurer scout out virtual realities in cityscapes and desktops solely in the mind of a computer. Within these virtual realities many things can be uncovered, much that looks like the real world, but to the real adventurer will come strange new symbolic abstractions: hypergraphs, concept maps and trees, and hirearchical structures of many kinds. Working out the engineering bugs from simulations that look realistic is the province of other books than this, This book converges on the issues connected with symbolic protrayals; how to make the mind's view public and real.

Unfortunately, we are only beginning the adventure. We are groping into a void with few guidelines to tether us, and few landmarks to direct our movements. As a result, there are many paths that thread into the wilderness without a known destination, and many more not yet taken. And yet, just as surely as Columbus and his colleagues knew that the world was round, we too know that the world of semantics has a structure; a network structure that we can simplify with hierarchies of nodes and links. Who will be the Columbus to chart this broad expanse of knowledge?

Every 500 years or so, it seems, something astounding happens to change the world. About 500 years ago two fundamental changes occurred to broaden our horizons. Johannes Gutenberg (ca. 1442) invented the moveable type that radically sped up the production of books, turning a lonely preoccupation into an industrial assembly line, while he ushered in the information explosion era. A few years later Christopher Columbus (October 12, 1492) began the age of global communication by discovering a continent and exploding forever a myth that held people's minds and aspirations in check.

Hypertext and hypermedia offer some of the excitement and liberation of both of these discoveries. More than BASIC or LOGO, or any other computer language it opens the new continent of computers to universal exploration. As did the printing press, so the computer too is revolutionizing something about books, through hypertext. As Figure 1.1 suggests, buttoning a hypertext link in a digital book can let you fly anywhere. The experience of zooming through text using computational relationships is addictive. For both reading and writing will never be the same, once hypertext and the old printing technologies are fully wedded.

Using this new technology is still very much an exploration in a new space of possibilities. Although there is general agreement that the possibilities are awe inspiring, exactly how this exploration will proceed and what will be found there, is a great unknown. Could Gutenberg and his associates possibly have foreseen that within one hundred years of his invention there would be over ten million books in press, and that every educated person would have his own library, not just a bible, but Plato and Aristotle and many new commentaries? Because books became so relatively inexpensive, everyone with a thirst for knowledge now had substantial access to the wisdom of the ages. It rapidly became easier to annotate the classics and to write extensive commentaries.

If a similar or analogous effect were to take place with computers, then it might be that the high price of education, its dependence on expensive libraries, and even more expensive research machinery, could be dramatically reduced in price and made available and accessible even to young children. Hypertext clearly offers a medium for new libraries. In fact, children should be able to begin with texts altered in some ways to make them more accessible, but with the complete and authentic forms lying just underneath, ready to be explored at a moments notice. Not only could this wealth of material be readily accessible to children, but they could grow up with these materials, annotating and personalizing them as the grow older and reuse the materials at different stages of their lives.

The idea of annotation and commentary is fundamentally altered by hypertext. One of the many things that hypertext does so well is to allow commentators to insert their own annotations right into the body of a text, without disturbing the original much more than placing an asterisk, a small icon, or any other barely noticeable mark in the text. But that note (like the quotes in Figure 1.1) then becomes a landmark and beacon that calls out to anyone who is interested that more relevant information is available on the topic at the click of a mouse or the touch of a finger. The annotation can then pop up and itself lead on into endless commentary upon commentary ( much like the Talmudic scholarship, the Upanishads, and the Apocrypha ).

Figure 1.1 The "quotes" around non-linear in the text below (if it was part of a real hypertext) would show the reader that he or she could zoom to another page in the text. That page could be related to "non-linear" as a topic in the organization of the whole text, or it could just be to an arbitrary spot, that the author of the hypertext chose on the spur of the moment, on a whim.

In a certain narrowly focused way, focused that is on the communication medium, the advance of science, especially during its early stages as the Renaissance was unfolding, the progress in clarification of basic ideas relevant for geometry, algebra, physics, and astronomy, can be viewed as an activity of commentary and clarification through extended annotation of earlier writings. Much of the effort of those earlier revolutionaries in science really was spent on rediscovering and absorbing even earlier Classical theories and bringing them back to life ( it was after all the Renaissance, not the Creation). Even as revolutionary theory as Copernican heliocentrism, for instance, owed much of its perspective to Ptolemy, and in fact remained grounded in the early texts, rather than dependent on new observations: a curious mixture of academic, rational theory and empirical generalization.

If hypertext had been available back then for theorists to share and compare their interpretations of ancient scholars, we would have today a very clear record of the development of their ideas, but how would it have affected the development of their ideas? And how will it affect modern science? Undoubtedly, they themselves would have been able to reflect more rationally on the slight changes each of them was making. Perhaps that would have given them a better understanding of the directions they were heading before they got there. Surely that is goal to strive for in modern day science, and potentially a powerful application for hypertext, especially for those sciences whose ideas are expressed only textually, or with only minimal use of other symbolizations or simulations.

Hypertext and Computer Access and Control

It is possible to create and manipulate computers efficiently and smoothly in very complex tasks simply by buttoning, selecting, and copying, and then sharing with others. More than any printed book, it provides a medium for communicating and sharing ideas and knowledge, in whatever form it most conveniently fits. If it is a picture, hypertext encourages the use of bitmaps and objects; if it is an activity or a task, hypertext encourages the use of a simple simulation; if its is a story, hypertext stimulates its telling in ways that engage and thrill, beyond mere words, beyond simply pictures and sound, into an unexplored dimension of hyperspatial interactivity.

What hypertext does so smoothly and seamlessly, is merge the control of the flow of events by computer and by user into one interactive stream. It does this by making use of all the power of the computer to enlarge the bandwidth of communication between you and the computer. It does this during the creation of hypertext and during its use. At both times, your personal computer becomes truly personal, closer to an extension of your knowledge and ideas than ever before.

Hypertext as a Knowledge Medium

Hypertext is a real knowledge medium. It exposes knowledge and begs for its refinement and analysis. Whenever you put a thought down on the computer, the power and culture of hypertext calls out to have it connected to another idea, to have it sequenced and ordered, both linearly and hierarchically. Hypertext exposes the hierarchical structure of ideas as networks of concepts and words and images. Ideas as associations connected into networks is a metaphor of ideas that has been around for quite a while, just as the notion of a round earth had been popular in scientific circles for centuries, but hypertext provides ways of making this metaphor concrete, just as Columbus' voyage of discovery did.

It is said that the scientific minds of Europe were largely in accord before Columbus' voyage that the earth was indeed round. There were many good reasons to believe so, from the earth's round shadow on the moon during lunar eclipses, to the disappearance of ships over the horizon. Instead, the quarrel that Columbus had with his contemporaries was largely over the exact shape and size of the earth: Columbus invariably underestimated the earth's size by a third. In general, Europe's geographers were about right and Columbus was ignorant and foolhardy in his underestimation. This had a pragmatic goal, of course, since an earth that is too large would make the voyage to India and China too long for anyone to survive it. Scurvy and disease or starvation would decimate the crew long before they could reach land. Fortunately for Columbus, America intervened and spared him certain death (although he went to his deathbed believing, apparently, that he had found India.)

If there is a lesson to be learned from this dramatic example, it may be that one can be right even though one is completely wrong. Columbus may in fact have known more than he shared. His hopes nevertheless inspired a new continent and shattered global ignorance on an absolutely fundamental point. So too, perhaps, hypertext will not be able to represent all knowledge conveniently as hierarchies and networks, but the effort to do so may provide the ultimate empirical demonstration of the true structure of ideas and concepts. And just as Gutenberg expanded the sharing of written ideas, hypertext expands the sharing of all networked ideas.

Shareware

Hypertext is radically liberating the power and worth of shareware. In the main, pieces of shareware are not scientific products of technical thinking, but inventions with new materials whose use demands some technical expertise. Unlike the inventions of the past, and most artifacts of today, shareware is flexibly aimed at a broad range of functions. Because shareware is implemented on that universal symbolic simulator, the computer, its use is not restricted to the uses contemplated by its creators. Because shareware produced with hypertext is modular and recombinable, its uses can be dynamically extended in any conceivable direction. Before hypertext, these extensions have not been easy enough for the vast majority of users to carry out. Imagine adapting a commercial wordprocessing system to your own idiosyncracies! Generally, you don't even have rights to the underlying uncompiled code without paying a special and exorbitant fee. And even if you had the source code, it is unlikely that you could make the changes: the code would be virtually incomprehensible to all except a handful of expert programmers. Shareware built out of modular hypertext no longer is so rigidly hidden from user alterations.

An alternative vision of software products is growing. Whether it is a meandering branch or the beginning of a new main channel is impossible to tell. It might be called coreware; a form of shareware that is open and encourages people not only to use it but to rip it apart, enjoy or bite off the irrelevant husk and replant the central seed in a new and fertile ground where it may be able to grow in new and unexpected glory. This kind of shareware may range from naive and simple to exceptionally sophisticated and complex. At its core lies the empowering use of environments based on hypertext and object oriented programming that demand fewer purely programming skills and much more sophisticated knowledge skills. These environments make up for their lack of speed and power with broader (but not yet universal) access. Yes, their strength lies in the very broad range of people capable and eager to accept the challenge of acquiring this minimalist expertise in order to share their inventiveness.

Cognitive Toolkits

The power of these environments come from the prodigious array of cognitive tools they provide. Unlike the programming languages of the past, these environments ask you not to learn their language; but to think, solve problems, design, and express your expertise creatively. They try to do all the easy stuff for you. They keep track of the lower code levels and demand of you the creation of interfaces, styles, flow charts, paths, and semantic connections. In a sense, these hypertext environments can already perform the basic skills of reading and writing, and arithmetic; they are missing the higher order cognitive skills of thinking and problem solving. But they are not entirely stupid. They have a modicum of intelligence. They act as cooperative agents and automatic helpers, taking the tedium out of software creation. They help to make software design challenging and interesting. The eventual success and growth of coreware as a medium for the exchange and growth of software products is predicated on the further evolution of these environments, the growth of electronic means of exchange such as bulletin boards and commercial networks, and fundamentally on the satisfaction and joy created by inventing and exchanging intellectual artifacts.

The proliferation of inventions in this area is staggering. Only a few years ago software was not treated with the intellectual property respect that it deserved: software was not patentable. The sheer force and volume of software inventions has gradually changed this situation dramatically. It is impossible to ignore the technical brilliance and native inventiveness of the myriads of people engaged in the exploration of this new space created by hypertext tools for inventing human artifacts.

Nowhere is there a community of intellectual competence and energy as ready and able to explore this new world as among teachers and trainers. For decades now, the promise of computational environments for instruction has been apparent to this community. Yet, the obstacles were always greater than the benefits. Now this has changed. The technology has finally matured enough to be accessible by the dedicated top ten percent of this population. This may be enough for the revolution to begin. As Ambron and Hooper (1990) expressed it, it is our hope that we can contribute meaningfully to the beginning of this educational transformation.

Teachers and Trainers as Professionals

Most of us think of ourselves as professionals in our work. But what exactly that means is not usually the concern of much reflection or analysis, apart from a dedication to upholding our integrity and maintaining a commendable level of understanding of the advancing knowledge in our field. Reading this book attests to your professional concern and dedication. But what about the knowledge and skills that we have? Let us try to examine them somewhat closer.

As professionals, I think we rightly claim to having privileged or even extraordinary knowledge about our field. Our sense of confidence and pride in our professional status derives from the large body of academic and research knowledge our field has accumulated and passed on to us through rigorous instruction and assessment of our skills. Whatever advanced degrees we possess provide an acknowledgement by our peers that we have acquired the minimal expertise to represent our profession.

We also provide a respected service to our communities. It is not enough that we have the knowledge, but we use it as wisely as we can in service of a larger group. When called on we offer our advice on those topics we believe are encompassed by our professional skills, as professionals, or make it clear that we are only acting as lay citizens for those areas outside our professional competence. Our clients are usually those people -- children, adults, organizations, other professionals -- who seek to learn and desire consultation on the technologies and techniques that will empower learning.

Yet, as professionals, we are often sensitive to a feeling of embarrassment and lack of confidence in the effectiveness of our own skills. All too often, we confess to an inability to explain to others how we arrived at certain decisions. They seemed to be intuitively right, but have no obvious justification in the academic body of knowledge we have acquired. All too often, our own insights have been just that: insights, rather than the deductive consequence of the technical skills our profession upholds. Especially as we try to evaluate new technologies for instruction it often seems to us that the old rules have been thrown out the window; that we have no clear-cut technical criteria to use, and that we are thrown back into a state of ignorance from which we have to flounder and thrash about to recover our self esteem. Where oh where, we then cry out, is the scientific bedrock that other more privileged professions (say, medicine or engineering) have at their disposal?

The comfort of technical rationality and the desire for scientific rigidity these views call for devalue the very same intellectual competence that is the cornerstone of professional ability (Schön, 1983). Simply because it is outside our competence at this time to describe in a scientifically rigorous way the cognitive skills and intellectual processes that a good professional trainer, teacher, or instructional designer possesses does not mean we can ignore them. Nor does it mean that these skills are any less effective. It would be nice to have scientific corroboration of the professional decisions we are constantly being forced to make. However, the fact that there is no corroboration or support is not because the decisions are wrong or unfounded or unjustifiable. In fact, it is much more reasonable to view our professional predicament as a failure of current science to deal with such complexity. Our professional judgments are simply too complex to be understood in the simple terms that modern science has to offer. At the moment, there is just no better alternative than to rely on professional judgment, and elevate it to its proper position as supreme adjudicator of our actions.

Of course, this offers no excuse for vegetating and remaining ignorant. Neither is this an argument for ignoring the developing scientific theories that offer the beginnings of an understanding of professional practice, and cognitive science offers just that. Never before have there been so many forces driving professional educators and trainers to adapt to rapidly changing conditions. The rapid development and integration of new technologies, both physical and intellectual, in the post-war "information society" have been unprecedented. Chief among these technologies, the computer has been particularly difficult for educators to accept (Becker, 1988). Not only did it call for skills and knowledge that educators normally do not possess, but until now, its chief use has been for learning, not teaching. Its main educational successes, LOGO and Computer - Based Instruction (CBI) of the drill and practice variety, have concentrated on student-computer interaction, rather than teacher-student-computer interaction. The new technologies of hypertext, hypermedia, cognitive science and artificial intelligence are changing that. Although they are part of the overwhelming complex of new technologies that a professional educator must understand if he has any hope of being competent, they are also becoming more directly connected to his skills and knowledge; and the barriers to implementing the rich power of these tools are consequently being lowered.

It is my belief that this is indeed happening; that the new technologies are creating environments in which instructional design can be carried out professionally: by those who can create instructional environments iteratively through a process of reflection in action, and those who can improve these environments through further experimentation in practice.

Reflection in Action

Although many of us would find it difficult to explain exactly why, we can offer excellent and detailed critiques of our own work or others'. How is it that one minute we can create our best effort, and then a few minutes later be dissatisfied with it and begin to revise by slashing components away with abandon?

Only a little thought will show that once having produced our first effort, the problem of design has been fundamentally changed; the whole context for problem solving has been altered. However we envisioned our design, the design itself speaks with clarity in ways our envisionments could not have. The design provides a specific instance to remind us of our past experiences and acts as a retrieval cue for relevant cases with a singularity of purpose that is far more specific than the plans and goals we imagined only a few minutes earlier. The product of our creative vision acts both to support our intellectual problem solving processes and to deny its errors.

Donald Schön in a truly enlightening book titled Educating the Reflective Practitioner has analyzed this process of reflective design as a kind of talkback. The design talks back to the designer. This is a useful, if somewhat unpsychological, perspective. Psychologically speaking, the design product is a kind of retrieval cue that reinstates all relevant past experiences and brings them to bear on the current situation. Some of these experiences cannot be fitted into the current design in the presence of others, without Procrustean changes. (Procrustes was a giant in Greek legend, who forced travellers to sleep in beds either too long or too short for them by arbitrarily stretching them or chopping off their legs!). Somehow, in ways that are not understood at all at the present, these experiences that do fit together are combined as a standard to judge and critique the current design. Mismatches between this experiential standard and the current product may be stronger in some components than in others, suggesting that those components be changed. It seems to make sense that the experiential standard, being an envisionment and not a real thing, is necessarily vague or ambiguous about many parts of the design. Only those components that are clearly discrepant stand out. But the changes may be as vague as the diversity in the set of experiences fit together into the standard themselves. These may yield very fuzzy suggestions for improvements in the design. And so this is why it is easier to critique than to design.

From this view, a professional is someone who has a large and conveniently accessible body of experience to bring to bear on a problem solving or design situation. Of course, design knowledge can easily be seen as the same as problem solving knowledge, if a design is held to be the instrumental solution of a complex, and poorly understood problem.

Applying this knowledge to coreware design means exploring the knowledge with the aid of these new tools. As a group, all of us are sharing in the recursive system of improving these tools for better design. Each design is a kind of experiment with the knowledge that we have, weeding out the inert knowledge (or finding ways to apply it) and using the specialized knowledge that we have as creatively as we can. Each experiment that works improves the tools for further design, in a recursive process of successive refinement that really knows no end. It shapes the future.

The Role of Research

Not only are we applying research results; better than that, we are carrying research into practice. The proper mix of research and practice has never been decided easily. All too often researchers are separated from practitioners by a large gulf of language, jargon, unshared knowledge, and attitudes. Practitioners are eager to apply researchers' knowledge, but don't know where to begin. Researchers are eager to supply useful knowledge, but the problems they can resolve never seem to have application and the practical problems seem too hard to deal with. By viewing knowledge as design (Perkins, 1988) research knowledge can be incremented simply by designing new systems: research can be practice.

difficulty in teasing out factors

James Coleman, in his High School and Beyond research (1982),
found that students achieved better in Catholic schools than in
public schools. At the time many questioned Coleman's methods,
claiming that he sampled too few Catholic schools. Later, his
results seemed to be more widely accepted and the effect was
attributed to stricter discipline in Catholic schools. Lately
(K. Bennett & M. LeCompte, How Schools Work, 1990), I note that
the effect is attributed to the practice of less tracking in
Catholic schools, with a consequent greater number of
academically oriented courses being taken by all students. How many more possible sources of this effect are there?

Unifying research and practice is a rather lofty goal, perhaps too lofty to be realized with today's technology. Yet, the advanced technology that is now available has such undisputed power for creating innovative learning and teaching designs that everyone, researchers and practitioners, can learn from inventing new applications and designs. It is all too clear that the simple design principles from paper - based instruction may be quite inadequate to deal with the potential complexities of these new hypermedia environments. It seems much more likely that only a teacher's personal classroom experiences will prove as useful as hypermedia experience itself. It seems indisputable that the only practical way to build new design principles is to experiment carefully and creatively with this new medium and record the results in new frameworks that can be shared by all.

Shareware and Design

At the core of creating these new instructional inventions and products to make use of all of these advanced technologies is a new view of instruction as design. Instructional design, a systematic and comprehensive process to create effective learning and teaching, has always been seen as an essential ingredient for dealing with the complexity of computer-based software. However, design all too often meant a begrudging subservience to painful detail and boring uniformity, rather than a creative exploitation of all the possibilities the medium of computer allied with video and even paper texts has to offer.

Design has been refocused by these new technologies in a kaleidoscopic way. To begin with, these empowering new environments for coreware creation encourage a more exploratory view of design: the emphasis has moved closer to the artistic end of the spectrum from the simply technical. Design involves a particular kind of expertise that knows not only how to plan. organize knowledge, and build; but especially how to listen to the construction as it grows. Listen to it as it talks back to the professional practitioner in its earliest rough cut forms, so that the design can be updated dynamically during the act of creation.

Letting the design talk back to the instructional designer encourages the growth of what Donald Schön has called the "reflective practitioner": a designer who is ready to take an iterative approach to creating his invention, even a recursive approach when needed. The reflective practitioner must be ready to engage in a dialogue with her construction, letting the design clarify the problem by ordering the issues that need to be resolved, and by posing the right kind of arguments at the relevant time.

Within the special field of instructional design, this kind of dialogue is best supported when the designer can act as both teacher and student within the same framework without needing to shift back and forth tediously and awkwardly in different environments. Acting directly as the student, the designer can get a direct impression of the content and structure of the coreware, letting it talk back openly and directly to her.

But this direct communication is not enough; it must be supported with structures that help the designer be reflective by helping her dissect the design into meaningful components, and by helping her navigate through the complexities of the design space itself which may have thousands of different alternative rules and heuristic that are appropriate at any particular aspect of the design.

A Framework for Design

Although we will return to this framework for design again later when its components can be better understood, we present a vision of the fundamental steps to creating a good instructional design now, as a kind of advanced organizer (Ausubel, 1978). An advanced organizer provides a skeletal overview or outline of the content of things to be learned. It offers a helpful scaffolding to hang new knowledge onto, so that even if at first the new knowledge does not cohere into a consistent structure, there is some hope that it eventually will.

It seems clear that we learn by drawing connections between all our fragments of knowledge. Hypertext may be able to make those connections concrete and explicit through its system of links; but at the moment, that is not a safe conclusion to draw. Connections between most facts are arbitrary at worst, and metaphorical or symbolic in most cases. Finding the connections and links among facts is often the hardest and only step that needs to be taken. A theory or framework offers one way of making these connections.

A very precise theory or framework for instructional design appears still to be a far off wish. Prescriptive statements for instructional design come from a cultural reworking of many generations of experience. In the case of hypermedia, we simply do not have an adequate body of experience, nor a sufficiently long period of reflection and digestion of that experience.

However, if we view instructional design generically as a case of a design profession, like architecture, engineering, or even graphics, then we can generate a practical framework that offers certain guidelines.

Guidelines for Instructional Design

o Analyze and decompose the knowledge and skills into bite sized chunks.
o Analyze the knowledge and skills of your audience.
o Determine your purpose and goals.
o Describe the structures and functions in terms of text and simulations.
o Pick a representative case or model.
o List your arguments pro and con; and the evidence for them.

There will be much more to say about all this in the chapters ahead, especially in chapter six, and more especially about the first part: analyzing knowledge into substantial but bite-sized components. There are many ways of creating such an analysis, but the most promising come from the new sciences of artificial intelligence. We will show how knowledge can be linked into hierarchies and networks on the basis of simple kinds of links and connections, and we will examine the tools for doing more than "rational analyses" from a desk. Instead the focus will be on empirical analyses using knowledge acquisition techniques pioneered in modern intelligent tutoring systems and expert systems. These create whole subsystems of knowledge that have been dubbed "problem spaces". Usually these are enormously complex, and hypertext combined with empirical and professional analyses offer our only hope for managing the complexity.

Constructionism and Instructional Design

In some of the most interesting applications of hypermedia so far in its early history as a technology, the use of instructional design has been liberated and offered to children as a powerful technique for learning (Harel and Papert, 1990). Continuing the trend defined by Logo and its history, Harel and Papert argue convincingly that access to the power of computers must be opened not just to teachers and trainers, but it must become so accessible that even children can use it to create instruction. Citing the common wisdom that no one really understands a topic until they are forced to teach it to someone else, Harel and Papert propose that Constructionist Tools should replace Instructionist Products, and that children rather than adults should be the designers and producers of multimedia instruction.

There are many radical implications of this constructionist perspective, but two seem to be of primary significance. The first is that current computer appliances in schools are far too underpowered, and in fact the approach to implementing computers in schools is wrongheaded. Rather than using cheap, bottom of the line technology, students should have powerful advanced workstations. It is only on these personal workstations that accessibility has been refined in a meaningful way for children, with object oriented , point and click interfaces and hypertext help systems, that make it truly possible for children to be creative and exploratory in these environments.

The second powerful implication of this constructionist view is that our most advanced models of knowledge representation for text and simulations must be embedded into the computer to support children's creativity in this system of instructional design. Given this support, hypermedia technology offers the possibility of a constructive alternative to drill and practice instruction: environments instead for knowledge exploration, reconstruction and representation, It is a vista of teaching and learning that is awesome to contemplate.

Hypertext

While hypertext has created many immediate opportunities, it also forces us to raise our standards of acceptable levels of instructional design and invites us to use ever more complex technologies. The flexibility and dynamics of hypermedia systems create new solutions, opportunities, and problems. This book will take an in depth look at the strengths and weaknesses of hypermedia, especially as they relate to instructional environments.

For a more rescent comprehensivee view of the role of text as a bsais for a Virtual University, please see Roy Rada's provocative Virtual Education Manifesto: Virtual Education Manifesto: Table Of Contents

Preface

Chapter 2