In 2003, I had a catastrophic equipment failure in my office. My working hard disk—including all of my manuscripts—and its backups were destroyed. Back then I never archived my projects, only backed them up, redundantly. I thought that was enough. I was mistaken. In referring to my earlier writings, I discovered that much of that writing holds up pretty well, so I’m reproducing it here for reference and the record. This article is from Macintosh Hypermedia Volume I, Reference Guide (Scott, Foresman and Company, 1990).
The third figure in the hypermedia historical triumvirate is a madman extraordinaire and one of the most brilliant minds of our time. How do you describe someone who carries around an encyclopedic knowledge base between his ears and simultaneously manages to maintain the spark of creativity? What are we to think of an individual who, when after almost 30 years of intense work finally receives adequate funding for his publicly accessible hypermedia repository, scribbles notes on his arm in purple marker during a press conference? How much stock should we put in the ideas of a computer visionary who generally refuses to use a computer? Sounds like my kind of guy. The caricatures are of Ted Nelson: the individual generally credited with coining the term hypertext and popularizing the concept by making it real to anyone who cared to immerse himself or herself in Nelson’s vast stores of rambling knowledge.
Nelson, influenced by Vannevar Bush, first used the term hypertext in the mid-1960s to describe a form of nonsequential writing. Most of his written works, most notably Computer Lib/Dream Machines and Literary Machines, have served to influence the current generation of hypermedia pioneers more than any other texts. If Bush was seen as a forward-thinker, Nelson has to be perceived as not of this planet.
His project of almost 30 years is Xanadu, a global information repository and network he refers to as the “magic place of literary memory.” Based on his concept of “universal hypertext,” Xanadu will consist of many thousands of nodes throughout the world, some of which will exist as fast-food franchiselike establishments Nelson refers to as “Silverstands.” When Xanadu becomes a reality—as it most assuredly will now that implementation funding has been acquired—many thousands of users will have simultaneous access to mountains of information, through which they will be able to create their own knowledge trails and endless document revisions. Of course, Nelson himself acknowledges that the name “Xanadu” is based on Coleridge’s unfinished poem, so there are no guarantees.
In the late 1960s, Nelson worked with Andries van Dam and a group of undergraduate students at Brown University to create the Hypertext Editing System, one of the first hypertext systems. The initial project was funded by IBM and was used for the Apollo space missions by National Aeronautics and Space Administration (NASA). The system, almost predictably, was not a commercial success and by 1970 Nelson was on to other projects. Andries van Dam and his students went on to create FRESS (an acronym for File Retrieval Editing SyStem) at Brown.
Hypertext by the book of Nelson
Ted Nelson, when referring to hypertext, means nonsequential writing, and by extension, nonsequential information retrieval and perusal. “Well, by ‘hypertext’ I mean nonsequential writing—text that branches and allows choices to the reader, best read at an interactive screen. As popularly conceived, this is a series of text chunks connected by links which offer the reader different pathways.” We also can extend the definition of hypertext to cover hypermedia by simply adding animation, sound, and full-motion video to the recipe. Nelson paints the hypertext concept with a very broad brush, initially using the layout of a magazine that employs sidebars and illustrations as a form of hypertext.
Quick to point out that hypertext could include sequential text within its realm, Nelson also referred to hypertext as “the most general form of writing,” for it was not limited by sequence and other external structures and conventions. Hypertext also would render a more enjoyable experience for the reader in that the reader would be able to choose a pathway to his or her own liking, rather than the strict one provided by the author in more pedestrian forms of communication. “Unrestricted by sequence, in hypertext we may create new forms of writing which better reflect the structure of what we are writing about, and readers, choosing a pathway, may follow their interests or current line of thought in a way heretofore considered impossible.”
Most writing is sequential, according to Nelson, because it grew out of speech making (as opposed to speaking) and because books are easier to read in a sequential manner. In the same breath he assures us, however, that the structure of ideas is not sequential, using a jumble of coat hangers as an apt illustration of the interconnectedness of our ideas. He also credits the concept of the footnote as a break from the sequential, but dismisses it because it cannot be extended. When explaining the hypertext concept to people, Nelson states that relatively few people grasp the concept right away. Most think it’s too technical or too philosophical. What they don’t understand is that writers write better if they don’t have to do it in a sequential manner and that readers read better if they don’t have to read sequentially. Nonsequential reading allows readers to form impressions and bounce around trying different tacks until they find the one that’s the most interesting or germane to their immediate task at hand. Hypertext allows for a totally arbitrary information structure, a structure that opens doors rather than slamming the doors shut.
Ted Nelson took pride in the power of words and was very much aware of the “spin” his word for his general concept would have once he named it. He admits looking for a loaded word like Bucky Fuller’s “dymaxion” which, according to Nelson, “condensed his philosophy of high-payoff engineering.” After quite a few miscues (“linktext,” “jumptext,” “nonsequential text,” etc.), Nelson knew he had the right word with hypertext. Later, he would reflect on the rightness of his hypernomenclature: “‘Hyper’ to most scientists and mathematicians means extended and generalized, as in hyperspace, hyperdimensional, hypercube, hypersphere, and even hyperchess .... Thus hypertext would clearly be the extended, generalized form of writing.” Nelson maintained a deaf ear to how the social scientists would react to his “hyper” moniker, including the protests of his father-in-law who was a child psychologist. He addressed this by explaining that his own quest for an understanding of interconnection had always led to chaos and that it was possible, even likely, that the two aspects of hyperness are two sides of the same coin—“deeply intertwingled” in Nelson Speak—and that disorder could be transformed into understanding by accurately mapping the interconnections.
Nelson foresaw that once we were liberated from the pestilent confines of the printed page, our writing would be helped to flow in a naturally interconnected manner. Additionally, a body of text could be authored without regard to a target market or “average” reader. Any level of detail could be achieved without concern for violating the supposed rules of general interest. Documents would be modeled after an onion rather than after a potato. Layers of detail could be peeled back and the reader could immerse himself or herself deeper and deeper into the writer’s work instead of the most cursory of treatments. Again, Nelson waxes eloquent: “I wanted everyone to see that we were going to the extended, generalized form of writing: no longer held to convenient sizes by printing and marketing considerations, no longer restricted to a single expository stream, no longer breaking the true interconnections of a subject to make a sequence (like branches snapped into sticks and put into a row).”
There are three immediate incentives for and related benefits of using hypertext: the allowance for preferences, motivation, and lower costs. Nelson correctly perceived that—especially in nonfiction works—readers have different preferences. In this book, for example, some of you may wish to read this section before reading anything else including the preface and introduction, while someone else may prefer to skip this section entirely. On the question of motivation, a reader is much more likely to grasp a subject if given the opportunity to explore freely rather than being “herded” through a text in a strict, predefined sequence. Finally, hypertext costs less to produce and distribute, therefore allowing the publication of more marginal works.
Nelson also was initially very careful to point out explicitly that computers had no essential involvement with the concept of hypertext, but that computers would most certainly be involved with the execution and development. Fully aware that hypertext would be perceived as “drastic and threatening” to the mainstream of society, Nelson took great pains to illustrate that the concept was fundamental to the vast corpus of civilization’s literature. Claiming that hypertext’s inherent nature of generality was a central theme recurring in our literature, Nelson bemoaned the fact that various computer systems were largely incompatible in a world of print where books and magazines were “at least unified and compatible.” He pointed out in an appropriately acerbic tone that one didn’t have to start a computer or initialize disks in order to open a magazine.
The generality inherent in hypertext, however, promised to tip the scales of the compatibility and ease-of-use issue by providing a sense of mass-customization in the universe of letters, a drastic and revolutionary idea indeed. “Customary writing chooses one expository sequence from among the possible myriad,” he explained, “hypertext allows many, all available to the reader.”
Nelson foresaw that hypertext would make writing easier, not more difficult. He noted that in sequential writing, the author must decide on the sequence which, in itself, was an unnecessary task of significant proportion. The hypertext author would be freed of the concern of sequence, and would be enabled to devote more time to the “interconnective structure” of the material. The writer’s flexibility would be enhanced greatly as sequential concerns give way to the simpler task of connecting text in a “searchable maze.”
Central to almost all of Ted Nelson’s work is the sense of pluralism and egalitarianism. Perhaps best exemplified by one of his own line drawings, reproduced here as Figure 1.2, on page 34, all of Nelson’s concepts and ideas seem to rotate around the notion of freedom and openness and the relations of many to many. Nelson feels very strongly that contributions from all users are important and form the basis of any information repository and that new pluralistic writing styles will develop as many authors add to a singular body of writing.
Patently understated, Nelson’s reasons for an egalitarian hypertext writing and reading system thread throughout his work: “In hypertext systems, there is a good reason to make the tools and access privileges of all users the same: the reader’s tool can be the same as the author’s tool. Thus hypertext may be intrinsically an egalitarian medium. (Provided, especially, that preexisting materials may be freely edited and incorporated in new objects without damaging the old.)”
Based on his insistence on the appropriateness and value of pluralistic documents, one of the main stumbling blocks for the Xanadu system as a workable solution for everyone remains the compatibility issue between types of computers, or, rather, the incompatibility between the various kinds of computers. “The initiatives and contributions of many people are assumed to be worthwhile. But there is at present no way to gather, and save, and publish, the many documents and scraps that people are writing on screens and sharing through an immense variety of incompatible systems.”
Interestingly, as much as Ted Nelson saw computers, especially personal computers, as appropriate hypermedia tools, he continually decried the concept of “computer literacy” as detrimental in that the issues taught to the noncomputer-literate are veiled in layers of unnecessary complication. Very eloquently he stated, “Nearly everything has to be fitted into oppressive and inane hierarchical structure and coded into other people’s conceptual frameworks, often seeming rigid and highly inappropriate to the user’s own concerns.” Nelson also took a firm stand against the traditional structure of the computer “file,” voicing a strong distaste for the “tyranny of the file” as illustrated by the file’s detachment from relationships and history that subsequently results in more confusion, not less.
A particularly common target of Ted Nelson’s venom was the early form of computer-aided instruction (CAI) that began to develop in the early 1960s. Nelson saw this as an attempted paternalism on the part of the schools at best and fascism at worst. “Though the student was implicitly at some position in a branching text complex, he or she would have no way to see it whole, no way to choose,” wrote Nelson in 1988, “the student’s only option was to answer questions, and these answers would implicitly make the next choice in a manner unseen.” This concept rested on the attempt to control and restrict users, where Nelson saw the promise inherent in freeing people to pursue their own interests, cross-references, and linkages. Always the pluralist, Nelson was adamant, “This was not the tradition of literature. This was not the tradition of free speech. It was the tradition of the most oppressive aspects of the bureaucratic educational system, dandied up to look scientific.”
From this distaste for oppression, Ted Nelson, the visionary, became something of Ted Nelson, the Protector of Our Rights. At the same time that he was singing the high praises of the hypermedia future, he also was sounding the warning bells of being ever-aware of possible encroachments on our freedom of speech.
Nelson’s broad-based goal, then, was a form of pluralistically general hypermedia, although he readily recognized that as the bandwidth of the component media grew, so did the potential for confusing disorder. Video, animation, and sound, while drastically increasing the bandwidth of the medium, also raised the potential for disaster and greater incompatibility was symptomatic of the situation. His proposed solution was elegantly simple: “To unify and organize in the right way, so as to clarify and simplify our computer and working lives, and indeed to bring literature, science, art, and civilization to new heights of understanding, through hypertext.” Nelson clearly perceived hypertext, and subsequently hypermedia, as a “framework of reunification” rather than just another obscure structure and duly noted Doug Engelbart’s initial concept that hypertext should be one piece, not haphazardly scattered about with bits here, there, and everywhere.
By proposing two styles of the organization of material within a hypertext document, Ted Nelson also demonstrated that hypertext would be much more useful for the reader than the more standard sequential forms of reading. He illustrated this succinctly by pointing out that when we read a work of nonfiction, we generally hop around from section to section to get the most information relevant to our current needs in the shortest possible amount of time.
Nelson referred to that style of hypertext organization that concerned itself with its possible effect on the reader—manifesting itself in a series of interlinked “planned presentations” the reader would navigate—as the “presentation and effect” style. At the core of a “presentation and effect” style of hypertext, the sequences would be designed for their look and feel and how they communicate their ideas to the reader.
The alternate hypertext style, which Nelson referred to as “lines of structure,” simply represented the organizational pattern of the subject matter. The effect on the reader of the material—in the lines of structure style—while taken into consideration, was not a major factor and was easier to implement for the author, “since the author is only concerned with analyzing and representing what the structure really is, and the reader is exploring the structure as he or she explores the text.”
Ted Nelson also was fully cognizant of the problem with reader orientation in a hypertext document. In my first attempt at electronically publishing hypertext documents for a general readership on a regular basis, one of the strongest complaints I heard from the majority of users was that it was too easy to get lost and too hard to determine what hadn’t already been read.
Nelson points out that in traditional paper publications the reader is given “incidental cues” as to his or her location in the material: “the thickness of a book, the recalled position of a paragraph on the left or right page, and whether it was at the bottom or the top.” He went on to propose that new cues must be developed that are equivalent to the cues we subconsciously employ when reading the more traditional forms of the printed word. Although initially he didn’t offer any ready-made solutions, several viable alternatives have managed to surface in the interim years between Nelson’s original proposal and commercial hypermedia systems.
Firmly believing that hypertext—with its inherent ability to present complex ideas companioned by their interconnections in the same documents—would advance the state of writing and learning, Nelson was fully aware of the potential of hypertext to address complex problems, which was Engelbart’s original supposition. Furthermore, Nelson envisioned taking hypertext a step further, to include the interconnections of many authors: “Hypertext can represent all the interconnections an author can think of; and compound hypertext can represent all the interconnections many authors can think of.”
Nelson referred to the interconnections of many authors as “compound hypertext” and began exploring the problem of revision-tracking as early as 1960. Referring to such arrangements of the interconnections of the multiple authors of a compound hypertext as problems of “intercomparison,” Ted Nelson correctly perceived a need on the part of the reader to be able to compare two alternative document structures on the same screen, side by side. At that point, Nelson set out to address the need of intercomparison and his work to that end is lumped together under the group heading of “thinkertoys.” The impetus for this work was clear: “Such intercomparison systems, I still believe, will become a vital aspect of our working lives—once they are easy to use.” Such devices and implementations are only now beginning to become available. Perhaps that’s why some of the best minds in the Macintosh community continue to work on video-display technology.
The Xanadu system, the embodiment of most of Ted Nelson’s work, was originally conceived as a solution to a very real problem while Nelson was a student at Harvard. He needed a note-keeping program that would serve as a repository for his thoughts. Xanadu, however, quickly grew in scope to encompass idea creation, thought organization, backtracking, alternative versions, automatic cross-references, text manipulation, and a complete electronic publishing system, including an automatic royalty-logging mechanism. Whew!
At the eye of the Xanadu hurricane is the concept that every document in the Xanadu repository has links to what it was drawn from and to those documents which draw on it. The linkages are electronic forms of footnotes except that the Xanadu flavor puts any would-be footnote into a separate window where it is readily accessible.
Nelson described a unique storage system for Xanadu, which he called “xanalogical storage,” that was based on a single repository that was shared across the system itself and at the same time organized in a myriad of different ways. Once again, the mass-customization of knowledge is at the base of most of Nelson’s concept. Originally designed as a vast dumping and storage ground for textual information, the Xanadu concept grew to encompass all interactive media as well and began to be seen more along the lines of a heritage preserve accessible to all community members. Because the knowledge base comprising Xanadu is a single shared pool, the contained materials can be continually rearranged to meet the needs of varied individuals without losing any of the prior organization structures. Sounding like a huge software program, in actuality, Xanadu is relatively small and relies on the cooperative processing power of all the nodes making up the network. Such an architecture further allows the system to be run by various individuals, under many localized ownerships, rather than being subjected to the potential of tyrannical rule by a single body.
Ted Nelson sees the current state of networking and electronic publishing as being in a state of small, mutually resentful bodies hopelessly embroiled: one against all others. He believes that a universal hypertext network would change that, supplying stored text and graphics on demand (again, mass-customization) from anywhere to anywhere else. Such a situation would, according to Nelson, render information “an elemental commodity, like water, telephone service, radio, and television.” Furthermore, Nelson holds that such a development would change the basic structure of information and would at last represent the true structure of information “with all its intrinsic complexity and controversy, and provide a universal archival standard worthy of our heritage and freedom and pluralism.”
Types of hypertext a la Ted Nelson
Chunk-style hypertext is best exemplified by the way we currently use footnotes and endnotes, additional text at the end of a section that illuminates a passage of text. Implemented as a footnote reference mark or symbol, when activated it would expand to display another chunk of text.
Collateral hypertext involves complex annotations and links created between two sections of text. It would include parallel text implemented in a manner similar to Nelson’s original parallel text specification.
Nelson’s original specification for displaying parallel text, the Parallel Textface, was designed to complement the Xanadu system and, in fact, some of the research for the Parallel Textface actually became the Xanadu system. At that point, the Parallel Textface became a “front-end function” in Nelson’s words. The Parallel Textface presents two portions of text, side by side, which are each windows on a larger body of text, with the windows scrolling together or separately, as appropriate. The window boundaries would be user-configurable by using a light pen to resize each screen element’s “pip” or handle. Pretty fascinating stuff for being ten years prior to the birth of Macintosh. As one window scrolled to show more of the underlying text, the other window, if dependent on the text displayed on the first window, also would scroll to maintain the visible link between the two text bodies. If there are no associated links, the dependent window simply stops scrolling.
Another Nelson proposal was the Qframe, similar to the windowing system we are familiar with on the Macintosh. The Qframe is a windowing environment for the creation, retrieval, and display of hypertext. All of the current windowing systems, including the Qframe and Macintosh, are based on Engelbart’s NLS system and got their boot into the real world from Xerox’s research at its Palo Alto Research Center (PARC) and, more specifically, PARC’s venture into the Smalltalk programming environment. In most windowing systems, relations to material in other windows are not displayed. Each window is displayed in and of itself with no connections to anything else, and therefore, no context. Qframes, on the other hand, overtly display all the interconnections between the various views on a body of text provided by the window convention. Each Qframe, however, maintains its independence from the others.
Nelson came up with the name Qframe because the frame of each window would contain information in the form of visual “cues” as to the contents and interconnections of the window. Immediate information concerning each Qframe’s interconnections and relations would be available at a glance. Qframes could be of either a hierarchical or hereditorial nature. A hierarchical Qframe would in effect serve as a browser to other connected windows, while a Qframe displaying heredity would show the interconnections themselves. Qframes also could be compressed so that more windows would be viewable on the screen at the same time.
Another Ted Nelson vision, the Walking Net, is a networked hypermedia system that is navigable through the use of a “walk-through” metaphor with branches, intersections, forked paths, and the like. The Walking Net is envisioned to be extremely simple to learn by users with the lowest levels of expertise. The user would be presented with a single one-dimensional choice at any given time, rather than being confronted with the myriad of choices inherent in any Xanadu information branch.
Stretchtext is a method of hypertext that enables the user to expand or stretch a small section of text into a much larger body that, in turn, expands into an even larger body of text. The metaphor employed is one of going successively deeper into a subject.
In the best case, chunk hypertext, collateral hypertext, and stretch hypertext would be combined in function to form a basic workable hypertext system, but would fall short of Nelson’s grand dream—universal hypertext—that would include everything on a given subject matter.
Conceived as a solution to the incompatibilities plaguing users of different computer systems, Ted Nelson’s concept of a universal hypertext was envisioned as the formation of an electronic literature that would set aside all the concerns of incompatible equipment and data structures. Where hypertext could be created and used by individuals, the call for greater data spaces useful by everyone, everywhere, all at once is addressed by the universal hypertext principle. In Nelson’s words, “that is the vision of universal hypertext: a world in which everything that is published becomes electronically available, in an ever-growing interconnected whole.”
The principle of a universal hypertext refers most directly to a new publishing system, and therefore travels back full circle to Ted Nelson’s Xanadu project. Once again, the concept is best thought of as the mass-customization of information. The key principle is that because everyone has access to everything, there is little need to have copies of much of anything in your physical possession. Another Nelson concept, “fragmentary publishing,” springs from this principle of everything everywhere all at once and nowhere in particular. It would be absurd to have to request a full copy of a title consisting of hundreds of thousands of words when all that is needed is to follow a link to the work’s third chapter. Mass-customization on a grand scale!
Even with a universal hypertext, however, Nelson continues to voice concern over the control issues: Who makes the links and who governs what kinds of links will be allowed? Nelson proposes that a universal hypertext is not synonymous with an open hypertext; an open hypertext is one in which anyone may contribute anything and establish his or her own connections. A universal hypertext would form a sort of hybrid, a system that is both open and universal.
So who gets to contribute? According to Nelson, “Some believe that participants must have special qualifications, that only certain persons anointed in some way are truly worthy of contributing; that information should be monopolized and that only some should be allowed to write and publish: that information should be a monopoly resource. They see the world of literature as something that has to be carefully controlled.” Nelson attributes such reasoning under a variety of misperceptions, generally centering on questions of professionalism. “But these are essentially styles of information monopoly, the assumption that Information Lords will decide what is to be available, and that the Information Peons (the rest of us) will have to accept what’s dished out.” Under any circumstances such a position is indefensible and not to be tolerated.
Nelson goes on to identify a variation on the information monopoly theme, that only authoritative versions should be allowed. Access should be available and granted only to approved information versions for the sake of stability and consistency. Nelson’s position is that all the works in all the disciplines should be available as well as all the comments made on them. Each work, each viewpoint, would be available as it was originally created—and clearly labeled as such—and no comments or corrections would be made on the virgin original. The links to other referenced material would be available to anyone who wanted them, but all the original works would remain intact.
Where the original material would be locked, free from modification (including correction), the links to other materials would be wide open. Anyone could link anything to anything else and anyone would be able to follow anyone else’s connections—in any direction. “This means not just the ability to follow connections in the direction they were originally made, but to follow them backwards as well.” Why would this be important? Individuals would want to know the origins of many documents as well as related or auxiliary documents. In order to see comments made on one’s own work, you would have to be able to trace the current document backward; it would give the corpus of massive works a sense of time and place.
The key issue here for Nelson, once again, is pluralism in the form of openness and freedom: “What I have been trying to communicate here is a sense of manifest destiny: we must recognize that the hypertext future is destined to be open, vast, free, and without restriction; with all participants and all links on a formally equal footing; in which intellectual property will not be an encumbrance but (like other property rights) a simple precondition, handled by a simple mechanism for automatic royalty payments and acknowledgment of origins. All this will bring a vast, new, practical, and intellectual freedom: What I call open hypertext publishing.”
As an example of the benefits of a universal and open hypertext system, Nelson explains that permissions to republish material (such as those that were required from various individuals for this book) would not need to be requested. Material could be republished simply by including the underlying work in the new document. The material itself would not be physically included in the new document, but, rather, a reference would be formed and linked to the original material.
An additional, perhaps more significant benefit would be the virtual elimination—by definition—of misquotes and taking things out of the original context. And royalties to the original author would be guaranteed by being automatically paid through a mechanism built into the system. A blanket inclusion and cross-reference permission would be implicit, by prior contractual agreement, before any individual work was allowed to be published on the system. Private documents would be allowed on the same system and could contain links to public documents, but the reverse case would not be allowed. Obviously, private documents would not be eligible for as broad a royalty arrangement.
Users requesting parts of public documents for inclusion in their new documents would not need to be aware of which documents the fragments came from, although the information would be readily obtainable. The only element the author requesting inclusion would be required to know would be the links used to access the document containing the fragments to be included.
The result would be an on-line electronic publishing system of vast proportions, easily transversed and accessible by minimally computer-literate individuals. Such a system would, over a relatively short period of time, serve to significantly enhance our collective body of literature. The most significant benefit to an open and universal hypertext system would be the seamless handling of the problem of multiple versions that, according to Nelson, requires a significantly advanced level of hypertext consciousness than we now generally enjoy.
Thinkertoys and super virtualities
Ted Nelson, along with Doug Engelbart and Vannevar Bush before him, perceived our planet’s greatest problems as involving “thinking and the visualization of complexity.” Similar in scope to Engelbart’s concept of “augmentation of intellect,” thinkertoys are more specific: a computer designed to help “envision complex alternatives.”
The crux of any thinkertoy is the ability of the device to allow things of varying levels of complexity to be intercompared and subsequently intercomprehended via their interconnections. Nelson gives very specific instances in which such devices would prove beneficial, ranging from alternative designs and theories to successive drafts of text documents to discrepancies in courtroom testimony. The underlying concept of the thinkertoy is that although the interconnections between vastly different problems appear to be vastly different, in actuality, they are more similar than dissimilar.
Significant differences between types of problems remain, however, leading Nelson to propose the most general of approaches to problem-solving including a technique he referred to as “collateration,” which is the “linking of materials into ‘collateral structures.’” (Nelson had previously referred to collateral structures as “zippered lists.”) Structures would be collateral if there are specific links between them, although the specific sequences of the connections may be different. “Collateration, then,” according to Nelson, “is the creation of such multiple and viewable links between any two data structures, in principle.”
To Nelson, the guiding principal of any computer system, regardless of its intended function, but extremely important in the case of a thinkertoy, is that any such system must be inherently, even disquietingly, easy to use. “If it is desirable that computer systems for simple-minded purposes be easy to use,” he said, “it is absolutely necessary that computer systems for complicated purposes be simple to use.” Patently obvious, yes, but exceptionally difficult to implement: Therein lies what many consider to be Ted Nelson’s greatest acumen. Power and apparent simplicity are not mutually exclusive in the eyes of Ted Nelson and he always has aimed at (and consistently achieved) the demystification of the various hypermedia technologies.
Nelson aims for simplicity to such an extent as to admonish systems that are more complicated than what he calls a “ten-minute system” (a system that can be learned by a novice and put to useful application in less than ten minutes) as almost useless. “I believe that interaction with computers can be at least ten times easier,” Nelson states, “ten times more powerful, ten times more vivid; and that these are issues not of hardware but of virtuality design,” he concludes.
In order for us to take the step toward the super virtualities, interactive systems with a level of power and simplicity that the original Macintosh only flirted with, however, we are going to have to let go of paper as an “ideal and as a security blanket,” and realize that “screen systems, without paper, will and must be the home of the mind in the future.” This is not a gushing statement to be taken lightly in a print publication, to be sure. Nelson points out several recent developments as enabling super virtualities, such as OWL International’s Guide hypertext system that is available for both Macintosh and the IBM PC, his own Zig-Zag hypertext system for individuals, Alan Kay’s Dynabook (a very powerful book-size computer that is easily affordable but alas, nonexistent), and Apple’s HyperCard software construction kit.
Hypertext and idea transfer
Hypermedia is, before and above anything else, a method of communication. Tantamount to any communication medium is the underlying concept of idea transferal. Hypermedia is a more appropriate communication medium for many endeavors, although it is not a communications panacea.
Ideas may be presented in any number of ways that are virtually unlimited for the articulate communicator: animations, various texts, graphics, sound, even with the face of the typography. Underlying the cosmetics of idea transfer—just as in the case of information systems—are the only guiding rules: simplicity and clarity.
Nelson also points out the value of a “socially neutral” information transfer, using as an example maps and some texts. I would hasten to take issue with the supposed neutrality of any communications medium. Like objectivity, there isn’t much, and what little there is can be indistinguishable from what isn’t, especially when such variables as time and place enter into the equation. We’re better off, generally, not to expect it and to be wary when we think we’ve found a neutral communications medium. Nelson argues that the best hypertext system would contain both socially neutral and socially active elements. I argue that this violates his own precept of no ruling body being in control of the medium. Who is to say what is neutral and what is active? A better position would be to consider everything as socially active.
The central problem, according to Nelson, is the widely held misperception of the writing process. Most nonwriters feel that the process of writing is to construct and subsequently fill in an outline. Nelson points to a National Safety Council pamphlet on how to survive a hotel fire as indicative of this general misperception. The pamphlet consists of “seventeen fine-print ‘step-by-step’ instructions [that] transcends imbecility… .”
Three procedures involved in the writing process sometimes occur simultaneously, according to Nelson. The first is what he refers to as the provisional development of ideas and points wherein the author forms his or her organizing ideas and relationships. The second procedure is sorting the points arrived at during the first process (yes, this can happen simultaneously with the first procedure). The final step is connecting the sequences using such mechanisms as cross-references.
Nelson called the structure of the organizing principles “arches” and defined them specifically as “final ironies, things to be led up to, themes, plots, concepts, principles, expository structures, organizing titles, overconcepts.” Nelson goes on to reveal that between the points used to elucidate the arches are transitions and that the main problem in writing as a method of idea transfer is that “overall structures you choose (systems of arches) may not link well to the points that have to be included among them; and that transitions between points don’t work out the way you want them to.” There are times when the points just don’t connect the “right way” or times when the transitions seem overly weak.
With the notion of idea transfer, especially if we take as a given the “activeness” rather than “neutrality” of the media itself, comes the possibility—even likelihood—of misrepresentation. Nelson bemoans the decline of diagrammatics (the ability to read and create charts and diagrams) in our society. He also goes on to illustrate the misrepresentation problem inherent in the concept of idea transfer using the broadcast journalism tactic of displaying an oversized arrow—either up or down—to represent the stock market’s activity for the day. No level of quantification is given in the visual cue; the market is either up or down, and we have no way of determining the size of the fluctuation. This problem is addressed exceptionally well in Edward Tufte’s The Visual Display of Quantitative Information.
Beyond idea transfer: Fantics
Ted Nelson’s formulated linguistics are seen by many as camouflage covering the weakness inherent in his ideas. Nothing could be further from the truth, although I’m sure Nelson would not take exception to being called a vaudevillian. Nelson as vaudevillian is best exemplified by his concept of fantics. In Nelson Speak, fantics is simply the showmanship of ideas. “I derive ‘fantics’ from the Greek words ‘phainein’ (show) and its derivative ‘phantastein’ (present to the eye or mind).” Well, that’s what he says, but I am not going to be convinced that he didn’t get it by combining fanatic antics. Some of the other words Nelson says he could have used for the concept were “teachotechnics,” “showmanshipnogogy,” “intelletronics,” “thoughtomation,” and “mediatronics.” Like I said, he is a vaudevillian.
In describing the underlying motives for the importance Nelson lies at the doorstep of the showmanship of ideas, he is quite explicit: “But I think it’s all show business. Penny arcades are the model for interactive computer systems, not classrooms or libraries or imaginary robot playmates.”
Call it what you will, but Nelson’s concept of fanatic antics to get ideas expressed and comprehended is right on the money. Contrary to closely held beliefs throughout most sectors of our society, the new media—and hypermedia in particular—will not require more and more technical specialization, but less.
All computers contain an inherent “learning curve,” that period of time required to learn how to use the system in question. New approaches to any problem have a learning curve attached to them: automatic teller machines, new cars, food processors, and computers. The beauty of the original Macintosh was its downsized learning curve. Anyone could be creating something useful on the machine in less than a half hour. Macintosh formed a new paradigm for powerful computing machinery. Software designed to run on Macintosh helped solidify this paradigm, and cries of “once you master any Macintosh program you have a great head start on most others” were heard throughout the land. How did this come to be? Aside from excellent design and the notion of “evangelizing” the product, it had to do with fanatic antics. People were actually enjoying working with computers—for the first time ever. So it’s not for nil that Nelson tells us, “I think that when the real media of the future arrive, the smallest child will know it right away (and perhaps first).”
Fanatic? You bet. With words such as “responsive computer display systems can, should, and will restructure and light up the mental life of mankind,” what would you call it?
Antics? Please. This is the man with the purple marker notes on his arms during a press conference, remember?
If the problems barricading the acceptance of hypermedia are not technical, what are they? “The fundamental issues are not technical. To understand this is basically a matter of media consciousness, not technical knowledge.” Central to this concept are four Nelsonisms surrounding the understanding and personalization of media.
“Anything can be said in any medium.” The techniques used to impart the idea are unpredictable, however, and cause most of the problems we associate with the myopia of most visual media.
“Transposability.” Related directly to the first Nelsonism, transposability speaks directly to the interrelatedness of all media. You can get your idea across in any number of ways, but generally one or more media are “better” or at least more appropriate for your task at hand.
“Big and small approaches.” Big ideas can be communicated any number of ways, and, generally, the smaller the approach the more effective the communication. There is always a more powerful way to transfer an idea, and the way to do so is not by massaging the content of the idea but by massaging the information structure used to transmit the idea.
“The word-picture continuum.” Writing and drawing—or, more specifically, diagramming—are a continuum, and where words are slow, graphic representation is lightning fast. Where graphics are generally inarticulate, words are fluent.
OK, fine, but what good are fantics in the real world? Well, consider that our intent is, first and foremost, to communicate an idea to either an individual or group of individuals. With that as our intent, then, we can say that fantics is concerned with vaudevillian showmanship as a presentation skill as well as specific techniques of presentation and the media themselves. From those three concerns we can distill that the basis for our endeavor is the design of the presentation system on both a conceptual and technical level.
In order to fully understand the concept and potential impact of fantics, according to Ted Nelson, we must understand that in the past, information structures, as the transporters of ideas, “sprang naturally from the nature of things.” This is no longer true, however, and “we must acknowledge that we are inventing presentational techniques in the new media and not merely transporting or transposing particular things into them because they seem right. The psychological constructs of man-machine systems may turn out to be largely arbitrary.”
Nobody knows and you can’t find out. You can, however, get in on the ground floor, roll your sleeves up, and put on your exploration shoes.
Thankfully, the human mind can draw unity out of vastly diverse concepts and ideas. This is one of the things we do best. Even more important, according to Nelson, is the fact that our perceptions also can be unified—he cites the example of our being able to “feel” with our fingernails even though we “know” our fingernails have no sensory nerves. “This principle of mental unification is what makes things come together, both literally and figuratively, in a fantic field.”
The best we can do as hypermedia authors and information developers is to help our readers/viewers make this “mental unification” in any way that we can. The most appropriate and powerful way to do this is through the use of an interface and control mechanisms that appear to be simplistic. To date, the best work on human interface issues remains Apple’s Human Interface Guidelines: The Apple Desktop Interface. Interface design issues are among the most important aspects of hypermedia production and will be covered only briefly in this volume. A more in-depth treatment is offered in HyperCard Development Tools, a companion to this work by the same author.
Hypermedia production entails many interface issues that must be addressed. Most center around control and navigation devices within the document itself and include, more specifically, text editing, searching, and retrieval issues; how the viewer/reader navigates from section to section without getting lost; how the viewer/reader backtracks to and through material that already has been browsed; and so on.
As important a consideration as the control devices themselves are the “real world” replicas chosen to represent the task at hand: Which guiding metaphor is employed and how do the other smaller supporting metaphors relate to the guiding metaphors. The guiding metaphor—the big picture—generally appears “all at once” and is usually arrived at first with the supporting metaphors (dials, buttons, slide controls, levers, knobs, and soon) following.
According to Nelson and most design experts, clarity and apparent simplicity is the easiest way to help the user from making mistakes, which, after the idea transfer, is our primary goal. A mistake free implementation of an appropriate metaphor in the form of the user interface will keep the reader/ viewer/contributor interested and at the same time won’t intimidate him or her. Instead it will invite, mostly by using techniques such as tension and compression to draw the user in.
For example, consider the original design metaphor for one of the first HyperCard stackware products, Mac TV. The opening screen consisted solely of a black background with a simple “on” button in a corner. I have yet to see a user who could resist clicking the button.
“Clear and simple systems are easier to learn, harder to forget, less likely to be screwed up by the user, and thus are more economical—getting more done for the resources put in.” The key to obtaining apparent simplicity in the interface and control mechanisms of a system lies in the generality of the thought behind the design. To the degree that the designer utilized generalized thought, the more simple the system will appear to the user.
The complete body of control devices, in addition to requiring painstaking thought concerning their individual design, also form a synergy that also must be addressed appropriately. This synergy resides mostly in the relationships and connections between the various screen elements. Again, this is best accomplished by thinking about the problem in the most general way possible. A single button out of place or of an inappropriate type could potentially affect the approachability and apparent simplicity of the entire interface. Nelson states the problem quite eloquently: “So the problem is to devise techniques which have elucidating value but do not cut connections or ties or other relationships you want to save.”
One of the most appropriate control devices available to the hypermedia designer is the map. A navigational map, readily available at all times, displaying where the reader is and where he or she can go (and ideally where he or she already has been) serves to orient the user and to provide a constant reassurance. Of course, in the future, we are likely to be able to call up a software “agent” that will be able to make an educated guess about where we would most likely wish to go based on where we are now, where we have been, and what the agent recognizes as our interests, but the concept of software agents is still a bit in the future. Even for us.
So there you have it. The history and underlying concepts of hypermedia in a nutshell. And now we can ask the $64 question with a gleam in our eye and a knowing wink: “What did Vannevar Bush, Preston Tucker, Doug Engelbart, and Ted Nelson all have in common?” May I have the envelope, please. They were all crackpots. They were all right. And all their ideas will last forever.
0 responses. Comments closed for this article.