"I recently took a speed reading course. I read "War and Peace" in ten minutes. It was about Russia. "
--- Woody Allen
"The feeling of speed and power soaring through knowledge in a hypertext can be exhilarating and addictive."
--- Anon.
"The philosophical mind unites where the pedant parts. He is convinced that in the senses all things are linked together, and in his desire for synthesis he cannot content himself with fragments.:
--- Johann von Schiller
" GO AWAY!"
--- Sign in a travel agency window.
Environments for Reading and Writing 90
Hypertext Expertise 91
Existing Models of Reading and Writing 91
Hypertext Hands-On!
Hyperties© versus Paper 95
Learning by Browsing or Retrieval 95
Narratives versus Hierarchies102
Intermedia 102
The Order of Writing 102
WE (Writing Environment) 103
Goals, Audience, Constraints 105
Storyspace 110
Sequencing and Linearizing 110
Editing and Rearranging 110
Genres of Reading and Writing 110
Novels 111
Technical Works 112
Journal Articles 112
Hypertext Reading and Writing 113
Golden Rules" of hypertext 113
Landmarks: Guidelines for simple sightlines 113
Individual nodes 114
Hypertext: Meaning 115
Heuristics for HAL (Hypertext Assisted Learning) 116
Environments for Reading and Writing
Although written script has been used by mankind for perhaps ten thousand years, and portable scrolls made from skin and papyrus have been available for about seven thousand years, books have only been made for less than two thousand years, and printed books, set with movable type, have been available for only five hundred years. In stark contrast with the centuries of experience with tesxt, Hypertext has barely reached adulthood from a single person's frame of reference.
Looking back, many of us are likely to think that when these innovations were first discovered they were immediately seized upon as solutions to obvious problems and opportunities of tremendous scope. However, this was probably not true in the past. Each innovation was confronted with the fact that existing expertise as stored by the craftsmen over the preceding generations was not immediately useful for the innovation, yet this expertise was the conventional source for producing and originating the innovation. For instance, it was not the clerics in monasteries who spread the technology of the printing press, even though they presumably continued to get many orders for books. Besides, the printing press did not suit buyers' taste for a good book. For instance, Rada (1990) reports that early printers used erasers and paint to make printed books appear more like handwritten ones.
Hypertext Expertise
How can there be any hypertext reading and writing expertise if hypertext is so young? In studies of professional composers and writers of fiction, it has been suggested that it takes at least a decade of continued practice to create expertise (Hayes, 1978), and that is in fields where expertise has already been defined over centuries of practice. Herbert Simon (Chase and Simon, 1973) has given a good account of the amount of time a high level of chess skill takes to acquire. He reasons convincingly that masters level chess players may have spent twenty thousand hours staring at chessboards. At forty hours per week, that would work out to about 500 weeks, or roughly ten years, solid. If each move takes about a minute (that is a gross overestimation to take into account reduplication in the opening moves of most games), the chess master would have seen 120,000 different board positions. It is possible that each one is retrieved or used somehow during a game in which it recurs. No one has had such experience working with hypertext. Most of the world does not have that kind of experience with reading plain text. Hypertext is far too young to have generated a profession of experts.
Existing Models of Reading and Writing
Since there is yet no evident set of hypertext experts, the only source of existing expertise is to examine how reading and writing is done for conventional print and media materials, and look at the existing prototypes of hypertext environments to support these. However, there are some limited direct experiences with hypertext writing and reading that may be fruitfully examined. While it can hardly be called "expert", it is at least and orderly systematic first attempt.
Hypertext Hands-On!
The first digital book offers one insight into hypertext writing.
DRAFT -- Do Not Quote or Attribute ---
(To appear in Educational Researcher)
With two 5 1/4 " floppy disks for IBM© PC compatibles.
-----------------------------------------------------------------
---------
I have just returned from an exciting voyage into hyperspace, and the vista was spectacular. I read this book on a computer, using a new style of digital text, called hypertext. I really feel that I want to share this experience, but obviously I am in a quandary. How do I introduce people to a novel experience and offer in a short space, and a fundamentally different format, a realistic perspective on this new stuff, without adding to the hype about hypertext? How can I be fair and impartial about something I value? How can I make predictions and offer advice about something that is obviously exciting, but still untried and unproven? How do I assess or compare something that is one-of-a-kind? How do I fairly review something whose authors have been heard to say: "Buy the book and throw it away?"
A Turning Point:
I agree with them. This is the first-ever hypertext book. Hypertext Hands-On! (HyHo!) comes with computer floppy disks that carry a Hyperties© version of the book, largely duplicating the content of the paper version. But the floppies provide the core of the hypertext experience, and the book can't do justice to that experience, nor does it really try. The Library of Congress has pronounced this the first digital book. A milestone! It is possible, or even likely, that this book will turn out to be the defining turning point of a new era of communication. As a representative symbol of a new medium of communication, this "book" points the way to the possibilities for education, although it is explicit about those possibilities only at a very general level. Still, any educational researcher worth his or her salt will find a wealth of ideas to intrigue and expand horizons here.
HyHo! is a non-technical introduction to a new branch of computer science that deals with digital text. If you have used a word processor you have some idea of what this new medium holds in store, but, be warned Your experiences in word processing will be misleading and provide only a glimmer of this new experience. If you digitized HyHo! and used your favorite word processor to read it, you would still be missing the fundamental hypertext effects. The paper version offers the same words, but reading the book without loading the floppies and running the software is a lot like kissing your sister. It's the software that provides an opportunity for insights into the potential uses of this new technology. The authors' suggestion of throwing away the book and only using the software is a realistic insistence on the value and importance of this direct experience.
The authors:
The authors provide an interesting combination of different perspectives. Ben Shneiderman is a distinguished professor of computer science at the University of Maryland. He is well - known for his work on human - computer interaction. He is also a pioneer in the development of hypertext and the originator of the field of software psychology. It is not surprising then, that his collaborator on HyHo! is an educational psychologist, Greg Kearsley, who is well-known for his talent for fluent and accessible (popularizing) writing. Together they have produced an engaging introduction to the field of digital textual databases.
Hypertext
The key to the new approach of hypertext hyperbooks is the point and click interface. Buttoning on a specially marked word (either with a mouse or with 'Hot keys') results in swift and direct traversal of the text. Sometimes this results in a new additional window with more text or graphics, and sometimes it results with complete replacement of the existing text. There are many hypertext systems currently available, and many more becoming available. It is a technology with a great deal of appeal, and it is being quickly incorporated into all kinds of software, particularly databases, online encyclopedias, and authoring environments for instructional software (courseware).
Each hypertext system has its own set of features and typical structure. HyHo! uses Hyperties©, a system originally conceived by Shneiderman and his colleagues. It is common to classify hypertext systems into two categories: those that scroll their text (holy scrollers), and those that present only one page or card at a time (card sharks). Hyperties© falls in the latter category, and HyHo! the hyperbook has been thoroughly affected by this constraint. It would have been a very different experience if Guide©, ToolBook©, or HyperCard 2© had been used. What exactly the differences would have been can only be a mater of speculation, however.
Another feature that would have dramatically altered this reading experience is an annotation facility. It is not possible to alter or annotate HyHo! as one reads it. There is no intrinsic reason for the absence of annotation capabilities (except perhaps a concern over copyright) and its absence frustrated my efforts at a coherent review on several occasions.
There are a number of experimental studies that might guide the users of HyHo!
Fortunately, because of Shneiderman's perennial insistence on empirical research, many of these studies have been carried out with Hyperties©.
Experimental Studies of Hyperties© that guide learning from HyHo!
Hyperties© has been under development since 1983. Hyperties© is designed specifically to be bullet-proof to users inexperienced with computers, and to provide rapid and pleasurable access to enormous amounts of information. It reserves authoring and writing capabilities in separate files for use by the creators of the material. Users need only the delivery engine, called the browser, to read the created material. They can then only experience the material, not annotate or edit it.
Hyperties© began as TIES ( The Interactive Encyclopedia System) with the implicit emphasis on being an online digital encyclopedia accessible by the (computer naive) public. Its primary function is as a database, and so it has very fast indexing and retrieval mechanisms while it remains very simple to operate. Basically it is possible to maneuver throughout the database with just two keys: the arrow and the enter key. It amintians a record of all the articles viewed and the order in which they were seen, so you can quickly backtrack through your reading experience. It is very fast, and provided its early users the first-ever direct experience of flying through a knowledge space.
Hyperties© provided the general public with their first genuine experience of hypertext in exhibits of the photographic art of David Seymour. Hyperties pioneered the use of touchscreen access to hypertext knowledge when in March 1988 an exhibit on "King Herod's Dream" opened at the Smithsonian in Washington, D. C. Unlike most museum exhibits, this one encouraged users to finger and touch the medium.
Hyperties© versus Paper
Many studies have shown that CGA-quality text on a computer screen slows reading speed by as much as 30 percent or more over paper. This disadvantage can affect finding facts in text quite substantially. In spite of this handicap, a study of using Hyperties on a substantial database of 106 articles of 50 to 2000 words each, found that as the complexity of the questions to be answered increased, Hyperties users began to find the right answers as fast as those using paper text. Full text information retrieval thus achieved parity with paper only in 1987. But even then, Hyperties users much preferred the assistance of the computer to doing it on paper, by a substantial margin. Even novices found it easy to use.
Learning by Browsing or Retrieval
The next step in improving on paper products for dealing with text, demands an improved understanding of how people find information and learn from these new environments. Tricia Jones(1989) highlighted the problem of hypertext design for retrieval and learning in her study of incidental learning during hypertext information retrieval. She used two groups of undergraduates to answer questions about environmental pollution. The Hyperties© articles were imported from Grolier's© Electronic Encyclopedia, reorganized, and rewritten. The main comparison was between students who used an index (and who did not browse) versus those who browsed the articles by using the embedded menu items to jump around (and who did not use the index).
Contrary to expectations, the students who were forced to use the index learned more incidental facts than the browsers. To a large part this was because they were misled more by the index than the browsers were by the embedded menus. Each embedded menu word was better defined and disambiguated by its surrounding context, whereas the index words were undefined. Organizing them into a hierarchical network might better have constrained their meaning. The browsers began from an introductory article that sometimes led them directly to the topic, or very quickly into the topic. The index, a flat alphabetical structure sometimes gave them no real clue about which article to read next, or worse still, sent them into a direction that led them away from their topic, into what Shneiderman has called, "hyperchaos".
What is so very clear from these studies is that the structure of the index and the structure of the links among articles needs to be vastly more refined with hypertext systems than with paper. In paper texts, the dominant access path is through the page number, and that is followed in a leisurely way by flipping pages. In hypertext systems, the page number is entirely hidden from the reader in indirect computations that whiz from one section or article to another. It is so easy to carom through this space, that without a constraining, organizing structure the reader will soon be retracing unrecognized steps in utter confusion.
The importance of alternative indexing schemes, such as hierarchical indexes with clear semantics, controlled retrieval queries, multiple windowing control, indexes for graphics and picture components, and the difficulty of converting from paper text to digital hypertext forms were all first confronted by Hyperties users. Consistent and efficient answers remain to be found.
Goals of HyHo! from the Perspective of Education
It appears to be the authors' clear intent to bring home the hypertext experience to as many people and as broad a cross section as possible.
Shneiderman and Kearsley both profess a strong interest in education and training, so it seems a fair action to judge the success of their hyperbook from the perspective of educators and trainers. Since this is the audience of this review, I will adopt this perspective as I approach the hyperbook, and provide something of a protocol of the result. I hope that this will make HyHo! more understandable to those who cannot peruse the digital version.
Browsing HyHo! for Education
Neither the paper nor the hyperbook version has any entry in the index that refers to education or training. The on-line version has the distinct advantage of having a search function. If we enter "education" as the search string, we get back seven "articles" (out of 248 in the hyperbook) containing the string. Each article is a one to four page entry in the database.
Most of these articles use the word "education" in a very incidental way. For instance, Greg Kearsley's vita shows he has a PhD in Educational Psychology. However, it is very easy to "flip" electronically through these articles by buttoning the titles or guess from each title which may be relevant. We will look more closely at two articles: one on other reviewers' comments, and one on "Instruction".
Other reviewers' comments (An Article).
Because of the novelty of this book, the authors have chosen to print a balanced and representative set of comments by early participant reviewers of this book. By "hot - keying" the title, we can bring up page 4 of this 7 page article of reviewers' comments. The word "education" is highlighted in the middle of the screen as the first word of a reviewer's judgement.
"Education in general, and computers in specific, have up until now rewarded linear, logical thinking. Hypertext, however, changes that by encouraging intuitive cognition."
I don't really know what that means. It seems to imply a powerful new functionality for hypertext; but whatever "intuitive cognition" is, it is pretty clear that education in general has rewarded much more than "linear, logical thinking"!
I seem to agree more with another review:
"A nice overview of hypertext that is highly readable. I think that designers/programmers can benefit by the clear and clever language and style, and the general reader can gain a good entry into what hypertext is all about."
But there is also a more critical review that makes me feel some agreement:
"To succeed , [HyHo!] should be both a good book and a good hyperbook. It is not. As a book, it is hobbled by having to fit into the hypertext style. As a program relegated to the least common denominator of [PC displays], it lacks what is good about books: illustrations, visible structure, and easy reading."
I think the criticisms are founded and apt, but as a hypertext HyHo! still works well despite its obvious limitations.
Instruction (An Article).
If we return to our search for "education" information, we can next examine another promising article on "Instruction". If we select the article and press the enter key once, a short preview or definition is provided for this article:
"Hypertext provides a new type of instructional medium with greater learner control and new cognitive strategies."
This is certainly an enticement to read the article to find out how learner control can be increased, and what new cognitive strategies are made available. Another press of the enter key jumps us swiftly to the article itself. A three-page article appears that briefly refers to four hypertext projects: CSILE ( an environment Scardamalia has been using over the last four years to explore how students can share and use knowledge in a student-generated database), Palenque (a video tour created by Bank Street in conjunction with the Voyage of the Mimi television series), Intermedia ( a hypertext system at Brown University for which an excellent video can be obtained from the Annenberg Foundation), and Project Perseus ( an analysis of extant ancient Greek writings, pottery, architecture, and culture). These are all notable early efforts, but it is somewhat surprising to find no reference to the extensive HyperCard© literature documented by Ambron and Hooper, (1988) that was readily available to the authors. Also there are no obvious links to learner control or learning strategies. The links that do show in this article are to Intermedia, Cognitive processes, databases, hypermedia, megabytes, CD-Rom, and tours.
These seven links provide an interesting menu to choose from. They emphasize directly the importance of the connection between hypertext and learner control. It is somehow up to us (the learners) to choose which link to follow on the basis of our interests and understanding. You, dear readers, cannot help me, but I decide that only cognitive processes and tours look relevant, so I take the first. To my surprise, this leads to a discussion of Vannevar Bush, the originator of many hypertext ideas in the 1940's. (I suppose he has been connected to cognitive processes because his conception of hypertext was based on a vision of an EXternal MEMory, he coined as Memex.) On the second page of the article there is a very relevant discussion of:
"The authoring challenge is to design the structure of the hypertext database to match the ways that a user might want to think about the topics. To do this successfully, considerable experience in teaching or explaining the subject matter is needed to know the many ways in which concepts might be related."
This clearly reflects on my surprise on finding Vannevar Bush as the main reference in the article on Cognitive Processes. In retrospect, there is nothing unusual about this, but it did not coincide with my expectations about cognitive processes and instruction. The preceding context had led me to expect something that the authors simply could not have foreseen in their arrangement of articles, indexes, and links. But instead, I was expecting some instructional work in the vein of the preceding four hypertext projects in education. The issue of student control is made complex by difficulties in providing navigational control, indexing, linking, and structuring the instructional material.
To compound the problem, we have now reached a dead end. There are no further interesting links with an educational context to follow. We must backtrack. A student in search of real instruction would find this a lot more frustrating than I do. But, in fact, there is a lot of potential meat here that is unexploited. The article is full of potential links that have not been made either by the authors or by the automatic features of the authoring system built into Hyperties©. On the first page of the article, they missed a terrific opportunity to get into the areas of hypertext that interest me particularly: knowledge representation, artificial intelligence (AI), expertext, and instructional design. A closer examination of the possible reasons for this omission reveals an important problem with indexing in general.
On the first page of the article, the authors gave a brief discussion of semantic networks; and this would have made a really fine introduction to the 5-page article on AI and knowledge representation. The link was never made. One plausible reason that it wasn't linked properly is that the word "networks" has already been reserved for links dealing with computer networks over telephone lines, and packet switching, and local area networks. As a result, semantic networks are not keywords or linked properly. The authors apparently used the term "hierarchical views" to replace semantic networks in other articles, isolating the cognitive processes article even more. This is an awkward phrase but it avoids the ambiguity of the multiple meanings (polysemy) of the word "networks". Since no existing indexing scheme deals easily or well with word ambiguities, the is a problem that is becoming more and more important
Continuing the Voyage
There are many more articles to look through, and my general judgment is that they are eminently worth perusing in this new framework. Follow the links where they may lead and judge for yourself what new uses this medium best suits. Hypertext for instruction appears to be very difficult to create well. Not only do you have to confront the problems of building orderly routes through complex knowledge spaces, but you constantly must keep in mind the framework and perspectives of many different readers with many far ranging concerns. One way of doing this perhaps is to create a basic set of materials and then follow many users through, keeping track of the paths they create. Most hypertext systems provide tools for this kind of book-keeping. Another direction for a solution is to use orderly networks of knowledge like those created by George Miller and his associates in their Wordnet system (Miller et. al., 1990) as an overarching indexing system to guide users through a knowledge space. Perhaps, too, the grandiose AI projects such as Lenat's CYC system (Lenat and Guha, 1990) will provide useful knowledge structures that will begin to impose order and coherence on these mushrooming and loosely ordered narrative structures.
This book and software provides a taste of a technology that is very young and immature, but growing faster than Jack's beanstalk. Already many different hypertext systems are proliferating, each with new and improved tools for instructional design and delivery. Entire curricula and environments are being built in virtually every subject domain: languages, history, and programming. It seems a safe bet to predict that a host of books on special purpose sub-topics of hypertext will rapidly appear ( Bolter, In press; Jonassen and Mandl, In press; Jonassen and Roebuck, 1989; McAleese, 1989; Nix and Spiro, 1990; Psotka, In press; Rada, In press). Yet, it seems very likely that none will offer the first-hand experience nor the dramatic sense of a voyage of discovery, opening a new frontier, that is offered by HyHo!, the first.
For a modern attempt to create hypertext interactive fiction, see: An experiment in Hypertext Fiction: The Intelligence
Bolter, J. D. (In press) The Writing Space: The Computer, Hypertext, and the History of Writing. Hillsdale, NJ: Lawrence Erlbaum, Associates.
Jones, T. (1989) Incidental learning during information retrieval: A hypertext experiment. In H. Mauser (Ed.). Proceedings of the International Conference on Computer-Assisted Learning at Dallas, TX. Berlin: Springer Verlag.
Jonassen, D. and Mandl, H. (Eds.) (In press). Designing Hypertext for learning. New York, NY: Springer Verlag.
Jonassen, D and Roebuck, Nancy (1989) HYPERTEXT/HYPERMEDIA Englewood Cliffs, NJ: Educational Technology Publications.
Lenat, Douglas B and Guha, R. V. (1990) Building Large Knowledge-Based Systems: Representation and Inference in the Cyc Project. Reading, Massachusetts: Addison-Wesley.
McAleese R. (1989) Hypertext: Theory and Practice. Norwood, N.J.: Ablex.
Miller, G. A., Beckwith, R., Fellbaum, C., Gross, D., & Miller, K. J. (In press). Introduction to WordNet: An On-line Lexical Database. International Journal of Lexicography, special edition.
Nix, D. and Spiro, R. (1990) Cognition, Education, and Multimedia; Exploring ideas in high technology. Hillsdale, NJ: Lawrence Erlbaum, Associates.
Psotka, J. Hypertext, AI, and Instruction. (In preparation).
Rada, Roy HYPERTEXT: Large - volume, collaborative, and intelligent. (In preparation).
Intermedia
George Landow of the English Department at Brown University discussed what he called the rhetoric of hypertext or in other words, the need for certain conventions for the links between nodes. HYPERTEXT preconditions the users to expect significant relations between files and a user becomes very disappointed if a link is not rewarding, and this disappointment may even lead to hostility towards the system.
Landow suggested links with labels or whose spatial proximity indicate their probably destination. Also, we could have menus of links related to some starting place. Another of Landow's suggestions for a style guide for HYPERTEXT structures is that most text nodes should be at most one screen full (bigger texts should be broken up). HyperCard has a set of icons which seem to have relatively well-defined meanings.
The Order of Writing
To order our overview of these systems, we will follow the conventional wisdom of how writing proceeds. We will also follow these steps through as they have been implemented in John Smith's Writing Environment (WE), since it has tried to be principled in the structure of the environment to reflect the cognitive processes of reading and writing:
WE (Writing Environment)
WE (Writing Environment) is a hypertext system (Smith et al, 1987) that is based upon a coherent model of the writing process. It provides four modes to aid writing. In addition to the traditional text and editing modes, WE provides network and tree modes to help the writer identify, organize and sequence the idea units.
At Hypertext '87 Davida Charney from Pennsylvania State University described a study of the reading strategies used by HYPERTEXT readers. These readers face the problem of loss of discourse cues. Traditional text which contains many such cues, ranging from genres (e.g. research paper versus. science fiction novel) over text-level schemas (e.g. the division of a research report into introduction, methods, results, conclusion, and references) to sequencing
These cues are lost when moving to a HYPERTEXT system which drops the reader in the middle of a new node in the same way no matter which node was the previous one. Also, in HYPERTEXT the burden of deciding when to read what has been moved from the writer to the reader even though structuring the material is one of the most important functions of an author.
On the basis of knowledge of how people read traditional texts, Charney conjectured that domain knowledge would have an important impact on reading performance in HYPERTEXT. Novice readers can be misled by superficial relationships and may stop reading too soon
Charney suggested the following ways in which HYPERTEXT designers may help readers: Design reading strategies which 1) depend on or reveal the underlying structures of the information space, 2) depend on precedence, and 3) have repetition or consistent patterns.
<<>>Kozma The Learning Tool
<<>>Scardamalia & Bereiter (CSILE)
<<>>Kintsch et al.
Gathering and Exploring Information
Powerful hypertext and hypermedia enviornments are now being used with
The design of these environments emerges from a rich tradition in
<
Organizing Themes and Clustering Related Materials
Inside Information
Stephen Glazier, a novelist, playwright, and editor conceived and developed a pyramidal database of defined words, structured in over 800 divisions. His original concept of a Word Menu has been commercialized by Microlytics as a Macintosh DA, called Inside Information.
As a hypertext dictionary, Inside Information provides a fabulous experience zooming through interesting topics, finding unusal words and short, pithy definitions. As in any web learning experience, one can proceed by jumping into the hierarchies anywhere. Using the "Find" option in the Inverse Dictionary, one can type in "NASA probe of uranus" and retrieve the word "Voyager" and the definition: " NASA probes to explore outer planets which flew by Jupeiter in 1979, Saturn in 1980, Uranus in 1986, and Neptune in 1989." Unlike a modern dictionary, it does not give you an etymology of the words nor the case of the word, nor even example sentences of its use. However, it does give you a sense of its proper context and place in contemporary thought, by placing the word in a modern category: Space Exploration & Rocketry. This can be almost as helpful as the definition in many instance, especially if the categorization is even more refined. In the example below, a fairly direct match is made to Voyager, without any real intelligence in the system or the searcher; quite by chance. Yet, this happens frequently enough to make one begin to suspect that natural lnaguage understanding and artificial intelligence may not be as necessary as many proclaim. Even more interisting results occur when the match finds several hits, some of which are surprising and unexpected. Of course, when there are too many hits ( 275 for "a word", for instance) it is time to use another probe.
The categories in Inside Information reflect modern, contemporary areas of general interest and importance. The category structure is not very deep, as Figure 4.2 indicates, usually four or five levels, before a long list of instances is encountered. Then the instances are displayed in simple alphabetical order in one long list. However, at any level of the hierarchy a fascinating hypertexty feature can easily be invoked: each instance can be double-clicked with the mouse to elicit a definition in the definition field, and to move the word or category into the center of the hierarchy. Moving through the lateral, coordinate terms can also be achieved by buttoning the scroll arrows under each field. The lateral and vertical excursions can be full of fun and meaning. It is a splendid way to find out new facts, jog your memory for old, explore the details of new understanding, and get a grand overview that may have been missing before.
Inside Information compares itself, and I think the comparison is very apt, to Robert Cawdrey, who in 1604 published his "Table Alphabeticall of Hard Words", the first English dictionary. In many ways, Inside Information captures the cream of the interesting relationships among categories and words, but it is far from being exhaustive in its detail, and may well be the first in a very long evolution of new knowledge documents that summarize mankind's exploration of the universe. In fact, it has some very clear affiliations with telephone yellow pages, in the topic structure and general interest of the categories. It may be that on-line yellow pages may indeed provide the level of detail that can be found in Inside Information. Just imagine, being able to explore the contents of department stores online. The organization might not be too far from the structure of information available in Inside Information
Jay Bolter, from the University of North Carolina, has created a unique hypertext environment, the Storyspace system. It is implemented on the Macintosh and is intended as a vehicle for creative writing of interactive fiction. Interactive fiction has existed for some time in the form of adventure games, and an online form is now found on the home PRODIGY system. Storyspace is designed to help writers control such stories.
The reader's experience of interactive fiction is dependent on how it is accessed. For example, the length of the individual episodes presented by the computer determines the rhythm of the story: For how long is the reader a conventional reader of sequential text compared to reaching branching points. Storyspace has two modes: A structure editor showing the links between episodes and a reader mode with a more limited view of just the text without structure.
Sequencing and Linearizing
Editing and Rearranging
Of course, there are many kinds of reading and writing too. These activities are very different for short thrillers, long historical novels, dense technical works, light humor, or condensed research reports in journal articles. We will examine three genres cursorily: the novel, technical writing, and journal reports; for the different perspectives they give on reading and wiriting as active cognitive processes.
A text may be seen as a serialized description of a more multidimensional vision in the author's mind. If it is successful and communicates to the reader, then parts of that multidimensional structure will grow in the reader's mind too, out of the apparent linearity of the text. The linearization of text is both a limitation and source of expressive power. The creative writers of the past and present have learned how to evoke complex and powerful structures within the constraints of printed text. Many others use fragments of text that recur throughout the book, forcing the recall of specific passages and deliberately creating non-linear connections, "leitmotifs" that link the text in complex ways. Interestingly, similar nonlinearities are used in opera and music.
Other kinds of nonlinearities in text are also frequent. Think of the complexities of plots and subplots in a modern murder mystery thriller. The red herrings scattered about create all kinds of complex structures. But these are relatively simple and plain compared the mental convolutions of Hamlet and Ophelia that Shakespeare created so masterfully. Calling a text linear, is as appropriate as calling Boston's streets straight.
As Lawler (personal communication) has pointed out in his analyses of Minsky's "The Society of Mind" (which he is converting to hypertext form) printed text may be viewed as a privileged but reduced presentation of the author's more complete vision. Hypertext is a way of superimposing different new dimensions to serialized text; a reconstruction, of sorts, of some of the multi-dimensionality of the original author's vision behind a text with the scholar's critical reflection on possible meanings of that vision and its sources. Yet, these annotations can be completely hidden and removed, unlike the footnotes and marginalia in paper text. The author's text, remains a privileged view.
Criticism and cooperation seem to be brought together in powerful new ways by hypertext. This is particularly true of the kind of fragmented book that Minsky wrote in "The Society of Mind" . Each page treats a topic from a unique perspective, yet, there are clear relationships among the different topics. Hypertext could prove enormously useful in tying multicolored patches together with common threads. One could also imagine this sort of technique applied to the aphorisms of Wittgenstein to create powerful new conjunctions and appositions of related ideas into a family of creative resemblances. It also applies to the world's first hypertext "book" , Shneiderman and Kearsley's Hypertext Hands On! The fragmentation of the many article's, and their ability to stand alone, separate from all the other articles in the boook, also makes them ripe for reconnection in ways the original authors might not have foreseen or intended, but might nevertheless find useful and in agreement with the spirit of their work.
But what if they disagreed with the new connections adn links that someone else imposed? Is it their right to prevent the formation and distribution of these links? How to cooperate and share in the growth of knowledge appears to be a new challenge to scholarship and learning. The very idea of plagiarism is raised by the specter of endless variants of the same material text, altered with only slight exegesis, being fobbed off as original work. How will hypertext writers and readers deal with this?
All structured text is basically hierarchical. In a novel, the book is the top level node, followed by chapters, sections, paragraphs, sentences, phrases, words, and characters. Of course one could continue this hierarchy down to atomic structures on each page. Several hypertext systems implement these kind of hierarchical structures directly in text. Guide, for instance replaces headings with the embedded text at a click of the mouse. And this embedding can go on to any depth. However, this simple hierarchy that allows people to read books (and reports and other documents) more easily isparticularly useful for journal articles.
Where are we to look for the beginning of authentic hypertext reading and writing tasks? As we have seen there is a kind of incipient hypertext in newspapers and journals. The structure is based on conventions and common practice rather than explicit hypertext features. Nevertheless, good journalists follow a common practice of writing in a fixed sequence with all the basic questions answered right at the start, and then separate sections later, elaborating the main points. Each paragraph or section is self-sustaining, so that if the editor decides, it can be removed painlessly without having to rewrite the entire story. This componential style has also been advocated for hypertext.
"Golden Rules" of hypertext
Ben Shneiderman's (1989) "Golden Rules" of hypertext focus cleanly on this componential division of text. His three golden rules for when to use hypertext effectively are that:
It is clear that he follows his own rules. The Hyperties system is a "card shark" system and so it forces its readers and writers to segment text into paragraph-sized chunks that fit cleanly onto a PC screen. This works well for a host of applications, particularly encyclopedias, museum exhibits, checklists of procedures, online help, and cross - referenced material. However, it is all too easy to pick a project ill-suited to hypertext, and even easier to create a hypertext system that is poorly designed. The danger than is that the reader will experience not the thrill of flying through high knowledge spaces, but the disorientation of hyperchaos when one is lost in hyperspace.
Design guidelines for hypertext reading and writing are still first approximations and crude implements, poorly fitted to a particular hypertext, not generally applicable to all hypertext systems. Traversing a book is a cleanly defined task. One starts at the beginning and stops at the end. There is a final certainty that everything in between has been covered.. Even stopping along the way, it is visually direct and easy to see that you are a certain fraction of the way through. The hypertext world has generally yet to build a counterpart for the well-defined vision of its content and shorelines. Sightlines in a hypertext are often confined to the local node. Overview browsers provide soem relief from claustrophobia, but only when the outline structure is simple enough to be grasped at a glance. Stop reading in the middle of a hypertext, and you may never find your way back to that spot again. Leaving bookmarks could easily be done computationally (Hanfling, 1989), but few systems offer this simple option. The most important landmark for any hypertext system is its overall structure, and that depends on understanding the knolwedge, classifying the key elements, and organizing them into clean hierarchies.
Providing an overview of a simple knowledge structure of the hypertext seems so easy to do. Yet, it remains a difficult and challenging task. It is not easy to analyse knowledge structures simply. The chapter on knowledge representation gives soem examples of the difficulties. Lenat and Guha's (1990) desctription of CYC gives all the hoary details. Nevertheless, it is important to keep in mind the kinds of structures that can easily be imposed on the hypertext
Individual nodes
Even when you reach a particular page or node of a hypertext, you may find yourself lost, especially if the node covers two or more connected pages in sequence. In Hyperties or Hypercard, for instance, the system accesses only the first page and some readers will not take notice of the "PAGE 2 OF 3" caption in the top right corner (see Figure 4.1). In that act of inattention, they will miss the rest of the information. Continuity within the hypertext should not rely on just one fragile cue. Scrolling text on a page does not have this drawback as severely, since the incompleteness of the text on a page is usually self-evident, and it is just as important to provide these "incompleteness" cues for non-scrolling systems.
Research on the relationship between node size and navigation is still fragmentary and
One of the best examples of a non-linear text, even today in the era of hypertext, is the modern newspaper. Look at any front page and you will see a large number of articles intertwined with references throughout the newspaper to continuing sections.
All structured text is basically hierarchical. In a novel, the book is the top level node, followed by chapters, sections, paragraphs, sentences, phrases, words, and characters. Of course one could continue this hierarchy down to atomic structures on each page. Several hypertext systems implement these kind of hierarchical structures directly in text.
Writing for HyperText:
Here are some guidelines for creating non - linear structure in text through a process of successive refinement that gradually shortens and summarizes the text.
1.) Build on Prerequisites: In a sequence of concepts delete those that are not prerequisites for understanding any of the following concepts.
2.) Use a Semantic Hierarchy: Replace concepts in a sequence with their superordinate concepts. (Be careful not to become too abstract, since this will become meaningless.) Keep in mind the lessons learned by children: stay with basic level concepts. For example, replace "Fido" with "dog", rather than "animal".
3.) Use Induction and Entailment: Induce a generalization to replace a series of examples ( vis. "bat" for "swung the stick at the ball"); or replace a series of parts with their whole (vis. "house" for "door, wall, roof, and basement"); or replace a set of components with their entailment (vis. "ran" for "splashed the dirty puddle onto the woman's clean dress").
5.) Use Abductive Explanation (Explanation Hierarchy): For each important fact or set of facts (or their generalization) create a model ( a kind of microworld or simulation) that includes the fact as a component and test its implications. The model must be supplied with important diagnostic criteria that includes symptoms, potential causes, tests, and manipulative actions. For example, for the fact "the sky is blue", create a miniworld in which there exists light, refraction, air, and water vapor. Allow the water vapor to be removed (among other tests) and see what happens.
Narratives versus Hierarchies
Books are complex narrative structures that are not readily converted into strict trees or hierarchies. Many people feel that hypertext will allow us to go beyond simple hierarchies in our instruction, and begin to approach the complexity of real problems in the complex ways they need. A prime proponent of this view is Rand Spiro (1988). <
Thes can be used to define link types for each node. For example, a "book" icon could stand for a "reference" link. By using these icons consistently, links could be "typed" .
John Smith (Smith, 1987) describes the funnelling process the author goes through in order to construct a linear document. For a hypertext the author is not obliged to cut out all the network links to form a strict hierarchy, and then lay out the hierarchy in a linear form. Instead, the information can be put into the hypertext in the same structure as it is in the author's head.
The Notes program (Neuwirth et al, 1987) helps writers record their own ideas, such as responses and inferences, and to review ideas from different perspectives. The ideas can then be sequenced to construct written text.
The most obvious use for hypertext is with reading and writing environments. Yet, somewhat unusually, these are not the dominant modes. In fact, there is really only one writing environment, that by John Smith (WE), although there are several good reading environments such as Guide and the Symbolics Document Examiner. Shneiderman's Hyperties can be seen as a reading environment, as can the Brown University group's Intermedia. Other more purely Notetaking systems have also been developed (e.g. Kozma's Learning Tool, Glynda Hull's reading instruction system in Hypercard.)
("there are three reasons forŠ, 1Š, 2Š,
3Š"), paragraphing and cohesive ties ("on the contraryŠ" etc.) showing how the previous relates to the next.
(before they have found all the necessary information).
<<>>Glynda Hull
Goals, Audience, Constraints
CSILE (Scardamalia)
fifth and sixth graders to give them a common database where they can
annotate, reconstruct, share, and reflect on their own intentional learning.
computer-assisted instruction and cognitive science of fostering learning
skills and strategies to help students learn how to learn, how to set goals,
and how to develop the metacognitive skills to control learning.
Observations of the first two years of students' using our environment
have led to design principles that might apply to the design of such
environments and more specialized software too. The overarching guideline
for designing such environments is to provide the kind of support for
learning that parents provide for young children who are learning language
and communication skills. Our approach has grown out of previous research
on the support requirements for instruction of written composition, in an
instructional approach we call procedural facilitation.
Some of our principles, which must be viewed as provisional and subject to
further observation and evaluation, include:
Make knowledge - construction activities overt;
Maintain attention on cognitive goals rather than task goals;
Encourage question asking by treating knowledge lacks in a positive way;
Provide process - relevant feedback;
Encourage learning strategies beyond rehearsal;
Encourage multiple perspectives on knowledge;
Support multiple knowledge representation techniques;
Encourage knowledge reuse and sharing for multiple purposes;
Support reflectivity and individual learning styles;
Facilitate transfer of knowledge across contexts; and
Give students more responsibility for their own and others' learning.
The presentation of networked information by computer has captivated many with its potential for learning (see Halff, 1988; Norman, 1973). Web teaching and learning can be easily produced by proceeding through the organized knowledge of a web structure. Knowledge related in semantic networks can be added to by proceeding in any direction from a particular node. Going to closely related knowledge might mean going up to a higher level of abstraction; or going down to a more concrete set of examples; or it might mean moving sideways to related items at the same level. Of course, we have been using level metaphorically without defining it too closely; although in the next chapter, we will go into much greater detail on the ways to order information in complex knowledge structures. For now let us take a broader view that is somewhat introductory to the next chapter, with an interesting example.
Figure 4.1 Using Inside Information to explore the meaning of unknown words, here, to look up " NASA probe of Uranus".
Figure 3.2 A Tree Display of the category hierarchy in Inside Information .
Inside Information is the first new classification system for language in many years. It has some characteristics of a dictionary and some of a thesaurus and something of an encyclopedia. Hypertext technologies are bringing all three forms together into a common format so that the syntactic and etymological definitions of a dictionary are being combined with the topical synonym structure of a thesaurus, and the depth of detail of an encyclopedia. Inside Information does not quite do all of these things, but it is a clear and dramatic step forward toward this structure. In a later chapter, we will examine George Miller's more theoretically rigorous and syntactic implementation of some very similar ideas.
Organizing Themes and Clustering Related Materials
Storyspace
Genres of Reading and Writing
Novels
What then is the realitonship between an author's text and a hypertext annotation and restructuring of the text? Unlike paper critiques and annotations, hypertext criticism can result in entirely novel and radically different reading experiences. Whole new trails and tours through the author's original manuscript could be created, resulting in radically different effects on the reader. In many ways, the hypertext annotation might even be more creative and a "better" reading experience than the original. Who will be given "credit" for this creativity? Criticism and cooperation seem to be brought together in powerful new ways by hypertext.
Technical Works
Journal Articles
Hypertext Reading and Writing
- there is a large body of information organized into numerous fragments,
- the fragments relate to each other, and
- the user needs only a small fraction at any time.
Landmarks: Guidelines for simple sightlines.
All structured text is basically hierarchical. In a novel, the book is the top level node, followed by chapters, sections, paragraphs, sentences, phrases, words, and characters. Of course one could continue this hierarchy down to atomic structures on each page. Several hypertext systems implement these kind of hierarchical structures directly in text.
Guide, for instance replaces headings with the embedded text at a click of the mouse. And this embedding can go on to any depth. However, this simple hierarchy that allows people to read books (and reports and other documents) more easily is not as important for hypertext structures as other more complicated, semantic structures
Hypertext: Meaning
Even if one took a very narrow view and said that books and magazines are written in a very straightforward and linear fashion on sequential pages (never mind what readers do with them when they read), this would still be fundamentally inaccurate. it is wrongheaded to think of traditional text as linear, because it is inherently a communication between people, and even the worst writer will take his potential audience into account while writing. These expectations of the reader all create non-linearities in the text. With better writers these non-linearities become more pronounced, and a host of nonlinear devices: repetitions, refrains, metaphors, and allusions proliferate on the pages.
Guide, for instance replaces headings with the embedded text at a click of the mouse. And this embedding can go on to any depth. However, this simple hierarchy that allows people to read books (and reports and other documents) more easily is not as important for hypertext structures as other more complicated, semantic structures
Heuristics for HAL (Hypertext Assisted Learning)
4.) Make HAL Bidirectional: If the text is too abstract, short, and general then invert the previous steps.