Hyperversity Apocalypse: A 1989 Prophesy

We know the word “apocalypse’ to mean the end of the world.  But it originally comes from the Greek for “revelation from above.”  Apocalypses were a set of prophesies about things to come, many about how the world would end. The most intriguing of the apocalypses are the so-called “Intertestamental Apocalypses” written sometime between the last work to be included in the canon of the Hebrew Bible (The Book of Daniel” – possibly second century BCE) and before the canonical works of the New Testament (possibly letters by Paul, 1 Thessalonians circa 50 CE).

A common literary device of these Intertestamenal Apocalypses was anachronism. The author would pretend he had found a manuscript written by an earlier figure from the Bible, say Abraham, or might even claim to be that author.  “Abraham”established his credentials by accurately “predicting” events that already happened and were well-known before proceeding to tell about events to come.

The following would be a nice bit of pretexting like these apocalypses, except I found it this week in a file that I recovered from an old external disk drive, into which I had dumped an old Zip disk, into which I had dumped files from old floppy disks and obsolete Macs. (I bought my first in 1984).


TO: Jim Meindl, Provost; Bill Jennings, Chair,

Committee on Computing at Rensselaer

FROM: David Porush, LL&C                         DATE: February 1, 1989

RE: Some input into your deliberations about the future of computing at Rensselaer

Attached you’ll find a long memorandum regarding a speculative design for what I call the “hyperversity.” Although the design I provide is couched in rather hyperbolic terms and implies a total overhauling of the university system, in fact it would be quite reasonable to implement the sort of hyperlibrary and hypertext-based courses I suggest without significantly disrupting the shape and function of the university as it presently stands. Indeed, it’sd probably healthier to look at the hy perlibrary as an enlightened complement to all pedagogical activities rather than as a substitute.

Toward the HyperVersity: A Design for the University in 2000 and After

Disciplinary and bureaucratic functions have slowly increased their dominion over the modern university, obscuring some of the more basic values in the education of young citizens. Rensselaer is extremely well-positioned to capitalize on its native talents (among faculty and support personnel) and its favorable disposition to new computing technologies to correct this imbalance by building and implementing a new method of university education, flexible enough to encounter and adapt to any challenge, but conservative of the important basic pedagogical ideals and functions of the university as they are presently constituted. This new university would be founded on hypertext techniques in the archiving and accessing of knowledgeessentially functions of the traditional library to build a hyper-library at the practical and theoretical center of university life. I call such a model for the postmodern university the HYPERVERSITY.

This memo offers

  • a theory of pedagogy underlying my design for the hyperversity;
  • why the goals of this theory now can and should be set as a central university function with the aid of state-of-the-art computing machinery, some programming, and the collaboration of the faculty;
  • a practical design of the hyperlibrary;
  • a portrait of how education will typically function in the hyperversity,


  • the redefined roles for professors, administrators, librarians, students, and researchers entailed in such a plan.

The interdisciplinarity of knowledge: Multiplying contexts for understanding

Whenever I look at the bank of steel file cabinets lining one wall of my office, I’m reminded of the modern university. Each file cabinet represents a different school, each school contains a different drawer or department, each drawer contains several sub-sections of faculty specialties or files, and each file, finally, contains the specialized and applied knowledge of individual people: professors, researchers, theorists, students. The rigid organization of the file cabinets is frequently overcome in the faculty/staff dining hall, in social arenas, in public events, in increasing administrative urging to create bridges between disciplines, in the formulation of new, interdisciplinary research programmes around new problems in industry, technology, or theory, and in interdisciplinary- and team-teaching experiments, and in the larger university culture of shared loyalties, histories, and expectations.

But more often, on a day to day basis, the university is a deeply conservative organizational system that tends to preserve its own cabinetry and filing systems, implicitly discouraging too much cross-talk or file-hopping.

The motif everywhere in this conventional university is specialization: institutional rewards, the allocation of resources and departmental accounting procedures, even the physical arrangement of offices make it difficult, if not impossible, to operate across the disciplines. Sometimes a new department will form, often around a new interdisciplinary interest (like computing, artificial intelligence, chaos theory is on the horizon), and an old one will lapse for lack of relevance, but even these changes tend to be glacially slow and do little to alter or challenge the fundamental structure of the file cabinets themselves.

For the student, this rigid organizational plan and the model of knowledge it implies becomes translated into rigid courses of studies with narrowly prescribed boundaries for completions of major and minor requirements. Finally, within courses themselves we almost invariably find what I call catechism masquerading as real education. As you know, a catechism is a method of instruction in which the student receives a manual or text summarizing the key points of a doctrine, often in the form of invariant sets of questions and answers which the student must memorize word for word. In catechistic instruction, the formal test, with numbered answers and quantifiably divvied dollops of information, becomes central in the student’s mind as a measure of what he has learned and also as a set of obstacles he must overcome in order to effect his rite of passage. That is why the courses that do not rely on quantifiable grades or present pre-packaged information as knowledge seem so exceptional. It is also why the courses that do are plagued by cyncicsm. In a recent conversation with a group of my students, some estimated cheating across the board in RPI courses to be as high as 70%.

The moral of the story is that any system that is so mechanical and simpleminded is also simply and mechanically (and cynically) turned against itself.

Knowledge is not a steel file cabinet

At bottom, what’s wrong with this institutional organization is that knowledge simply does not work like a steel file cabinet: phenomena are not disciplinary.  Take a bridge, for instance. Is a bridge the expression of an designer’s blueprint or is it a transportation device? Is it an architectural creation or an economic one? Is it a social fact, linking two communities, or is it an ecological one, disturbing a riverbed and the river’s flow? Is it an historical fact, layered with accretions of symbolic associations, or is it a mere mundane fact, a convenience for those who use it? Quite obviously, a bridge is all these things and more. It is almost anything you make of it, and what you make of it depends on your point of view, what questions you ask of it, what’s motivating you to look at it and question it in the first place.

A colleague in Mechanical Engineering, Gary Gabrielle, asks his students to design a real-world device in order to complete his course. In one instance, his class works on building and designing a motorized wheelchair for children with muscular dystrophy. The student’s first goal is to define a set of desirable parameters. But before they do that, they must meet with the children themselves, interview those therapists who work with them, and try (though Gary doesn’t explicitly pose the problem to them in this way) to imagine themselves as afflicted with the same disease, forced to deal with conventional wheelchairs . Only then do some crucial aspects of the problem come into focus. The vehicle must be shorter than conventional wheelchairs. It should look more like a toy, in contrast to the forbiddingly technical look of conventional wheelchairs. It needs a low center of gravity for safety. It must be designed with adjustable controls so that the physically challenged children can learn how to use it. In one most poignant definition of the complexity of this problem, Gary talks about how at first the joystick controls must have self-limiting responses because if the chair over-responds to a motion and scares the user into reacting too quickly, the disease has the idiosyncratic effect of making the user freeze up in one total body seizure, which might in turn push the vehicle to respond dangerously, not to mention embarrass the user and make learning how to use the wheelchair even more daunting.

In short, what looks like a tough but manageable problem in mechanical engineering soon unfolds into a much tougher, much more complex, almost unmanageable interrelation of contextual problems that are physical, psychological, social, experiential, environmental, aestheticÑ or in other words, human. In addition, these students work in teams, and must figure out how to manage not only the contexts for their project, but each other: what if one student is lazy or ill-qualified? What if another likes to boss his teammates around? What if everything hinges on the completion of a minor task and the student who was supposed to do it falls ill?

By no means could a steel-cabinet approach succeed in Gabrielle’s course or solve these problems easily. While specialized knowledge is absolutely necessary, it is not sufficient: : most of the time the real world poses us problems where the solution relies on getting a three-dimensional grasp of the slippery human contexts.

The failure of the file cabinet approach is most easily and painfully seen in the careers of our students. As Jim Meindl is fond of saying, RPI graduates who go to work for larger corporations have little difficulty getting their technical (first and third) promotions but a lot of difficulty getting their managerial (second and fourth) promotions. And whether or not statistics actually bear this out, there is an important perception there about the skills, our students get Ñ or don’t get Ñ from Rensselaer. A manager is invariably the problem setter, the one who frames the contexts, the one who sees the bigger, more indefinable picture and asks the larger questions and knows how to express them succinctly for others.

If you accept my premise that most tasks, even apparently mechanical design tasks, are more like the hyper-wheelchair design problem Gabrielle’s class faces than like solving a tough partial differential equation, then you might agree that there is a fundamental mismatch between the steel cabinet and the world out there. Furthermore, this mismatch makes the university into a place that prepackages knowledge into easily consumable and digestible bits, grossly oversimplifying how the world works. Adult teachers and administrators with superior experience aren’t easily deluded into accepting this oversimplification, but I’m afraid that our students often are. It is very hard to persuade a student who is told he ought to declare a major at the age of 17 or 18 that the world isn’t a specialized place. Similarly, it’s very hard to persuade a student who is trained to give rote answers to pre-packaged questions that the world isn’t similarly catechistic.

Multiplying contexts for thinking

Opposed to catechistic models of learning are the exegetical and dialectical models. In exegesis, knowledge is viewed as something you acquire by inspecting and interpreting problematical, open-ended events (texts) for which there are no clear solutions. For instance, where catechism relies on student’s abilities to memorize the “right” answers, exegesis requires that students uncover hidden information, decipher systems of signs, piece together clues to a puzzle that might not have one proper arrangement, choose the right tools or formulas to “open the box,” and ask the right questions. Where catechism excludes alternatives, exegesis multiplies alternatives. Catechism is woefully limited, exegesis marvelously unlimited. In catechism, there is only one, higher final authority. Exegesis teaches the searcher to authorize (and be responsible for) his or her own answer. The dialectical element in learning in emphasized: students inquire of the field of study, engage it in a dialogue. The field of study – in the person of an instructor or professor – in turn, asks questions meant to spur the student to arrive at his or her own answer. This model of teaching passes the litmus tests that an Rensselaer educational fails: how easy is it to cheat the system?

The problem then, is how to get students to begin, even at the earliest stages of their college careers, to learn how to multiply contexts for their thinking, take responsibility for their solutions to problems, and appreciate the value and necessity for their future careers of doing so. Certainly, continuing to divide “knowledge” up into narrower and narrower foci, or simply multiplying file folders, or interrogating students by “plug and chug” catechisms not the solution, for even in its intrinsic structure, such an organization communicates the wrong message.

The solution, I would maintain, is within our grasp, and is especially important in the context of Rensselaer’s deliberations over the future of computing here.

Hypertext as curricular model

If we want to dispense with the steel file cabinet metaphor and the catechistic interaction between professor and student it too often requires, then what should take its place?

I would like to offer a new metaphor: the hypertext programs now popular in the microcomputer world.

Hypertext is, quite simply, a new, non-linear concept of accessing information made possible by the raw power of the contemporary computer and some imaginative programming. Hypertext has been expressed first as several high level programming languages, including the popular HyperCard (for the Macintosh) which make it extremely simple for the software designer to multiply contexts or different points of entry or ways of viewing the information held in large, complex databases. In its purest form, a hypertext system takes a finite but large collection of data, even data held in different media (video, audio, text) and enables any number of routes to cross-correlate – or build bridges between – aspects of those data. In common parlance, such multiplication of opportunities to see connections creates a “knowledge environment” and some hypertext programmers style themselves “knowledge architects”

SCENARIO 1: In a project already underway at Rensselaer, my colleague Pat Search and several associates are building a hypertext environment for the Massachusetts Museum of Contemporary Art. According to their description, a hypertext system will make the following scenario possible:

A visitor to the museum with no prior training in the use of computers is intrigued by a Robert Morris sculpture in one of the galleries. Wishing to find out more about this artist’s wrok, he walks over to a large computer monitor on a nearby gallery wall, calls up information menu, and chooses to see a series of images of Morris’ major works, many of which are in private collections abroad.

Stopping at an image that especially interests him, the visitor presses a key to zoom in on a detail, then requests and receives a 360-degree video view of the entire piece, as well as a (written) review of the work that appeared in a leading art magazine. In the middle of reading the review, which includes a description of the artist’s creative process, the visitor calls up a video of Morris making one of his sculptures. [From press release, 1/4/89]

Hypertext permits the student to continue this process almost ad infinitum. In the hyperversity, the only limit on the student’s field of inquiry would be his or her own curiosity ( or the problem set him) and the size of the database on-line. The student could call up a history of Morris’ development, an account of artists who influenced Morris and pictures and reviews of their work. Given a large enough resource (database) the student of the work could go even further. He could examine a history of sculpture in America, or find out more about the material used by the sculptor and its special properties, down to its mechanical and thermal coefficients. He could call up other videos, including tapes of people viewing the sculpture and artists creating other sculptures. He could discover relations between engineering and sculpting. In terms of his goals in a particular course, a student could request to see final reports, tests or work performed by students who have taken the course in the past. He could look, if the instructor was willing to offer the information, at the grading history of the instructor or at the instructor’s professional resume. At any point, the student could

(A) ask for a bibliography for further reference, and either call up entries in that bibliography or simply print it out for future reference; – or –

(B) ask for a printout of selected textual material or simple diagrams or images; -or –

(C) Ask the computer to suggest related topics and keywords (GUIDED SEARCH).   -or-

(D) Ask for definition or explanation of ambiguous words or for entries on particular names, things, places, dates, etc.

SCENARIO 2: Students in the Shakespeare Project at Stanford University can access three or four differnt performances on videotape of several of Shakespeare’s plays. They can watch the video as the text of the dialogue appears on the screen, and at any point stop the video ask for commentary on the text, access an archive of thousands of photographs and drawings of stage sets, costumes, and props. Or they can switch to the performance of the same scene by an entirely different company (for instnace, juxtaposing Nigel Williamson’s famous Hamlet with a recent and more sombre Russian performance. Furthermore, they can animate their own version, using stock animation characters that they can manipulate on the video screen, dress up in different costumes, and place in fornt of an array of different stage sets. (See “The Inter-Play’s the Thing,” MACUSER March 1989, pp. 108-114)

Now your reaction to this might be, “Fine, such a system works well for the fuzzy sets of liberal arts and social science courses, but what about technical courses where hard, invariant facts and formula are crucial to a student’s success?” Well, ignoring for a moment the discussion we might have over whether staging a production of Shakespeare is a highly technical enterprise or a mere liberal arts activity, let’s see how hypertext might apply to a more conventional Rensselaer course. Here the answer lies in the virtually infinite adaptability of the hypertext system.

SCENARIO 3: Imagine Calculus, first week. A student wants to know the meaning of the integral sign. She calls up a simple definition. But could also call up history of the choice of the letter ‘d’  by Liebniz, the history of Newton’s development of the fluxions (including animations of the exact physical problems they tried to solve), perhaps set in terms of a problem for the student to solve an interactive situation in which the wrong answers would also be animated. The student could also witness and manipulate animations of the increment under the curve, limits, changing the curves according to different algebraic formulas, velocity and acceleration, and alterations in the curve upon derivation and integration. Furthermore, illustrations of several real-world problems or applied calculus coudl accompany any theoretical discussion.

Technical skills are already taught in interactive hypertextual mode, as the famous diagnostic programs, with videodiscs, used in some teaching hospitals. In fact, the distinction between vocational-technical training, on the one hand, and liberal arts learning on the other, actually should becoming harder to make in the hyperversity. The positive feedback loop – the absolute symbiosis – between culture and technology/science becomes quite apparent if one follows the thread of any problem through all its labyrinthine twists and associations.

The advantages of a hypertext system for learning should be immediately apparent:

  • It permits students to proceed at their own pace;
  • It’s self-guided, allowing students to follow almost any avenue;
  • It’s knowledge-oriented rather than task-oriented; it spurs students curiosity first and gives a more realistic sense of the inexhaustibility of knowledge;
  • It’s multi-sensory, stimulating the brain through 3-d visualization, animation, video records of actual authors, events, objects in motion, and sound and color, through dramatization and dialogue, as well as through the linearity of written texts and taped lectures;
  • It’s open-ended and permits student creativity at all points rather than pre-packaged and finite;
  • It portrays the inherent inter-connectedness of disparate fields of research and different realms of learning;
  •  It contextualizes information, giving meaning to data by showing their place in the larger scheme of things;
  • It asks professors to redefine their own goals in teaching. Professors now become not so much conduits for bare facts and formulas, but are freed to become problem setters, resources for human knowledge that can’t be “uploaded” to the database, models for ways of thinking and styles of problem solving, rather than administrators and accountants of problem sets.
  • It brings students into more intimate contact with the real world, helping to break down the unrealistic conjurings a university education can sometimes inspire

But in my mind, the most important consequence of such a design is that

  • It puts the entire information resource of the university – and soon enough, I’m sure, all libraries that are digitized  – at the disposal of every student for the investigation of every problem.

The pedagogical problems with such a system, like most sophisticated tools, are a result of its very strengths. For immature researchers (our students) it can create the Alice-in-Wonderland Effect, bombarding the hapless student with an overwhelming amount of new information, unfamiliar viewpoints and dizzying contexts. However, in the structure of well-defined course goals (particular problems to be solved, a limited time in which courses to occur), the problem of infinite horizons can be controlled and managed. Professors can also easily design their own routes in the system which will create “more favorable pathways” for inquiry without blocking others.

Another problem is that it would be very easy to build inert hypertextual courses, recapitulating, in effect, the problems of our present system in a glitzier, fancier, more high-tech format: students could be made even more passive by relying on hypertext sessions like they do on TV (the “FEED ME” effect), thus losing the inter-active stimulation advantage such systems have. In short, we’d be left with a very expensive hypercatechism.

The way to avoid such a scenario is to embed deeply within the values and technology of the hypertextual system a sense of exegesis.

How to build a hyperversity

First and foremost would be to expand our understanding of the library as a central resource. With their skills at knowledge organization and resource management, librarians would work closely with professors to upload and archive everything relevant to courses taught here. They would then organize favored pathways for directing the researcher (student, user) from place to place, especially linked to particular courses. Pat Molholt informs me that there are already steps underway to build a hypertextual sort of administrative shell to help students select courses. She also has told me about plans to build a primitive hypertextual search procedure employing a sort of “shelf” search in on-line requests, in which a user could not only ask for titles with a keyword, but for titles one would find next to keyword titles on the physical shelves of the library. (This of course relies on the underlying Library of Congress system for organizing knowledge, but it is a good prototypical start of hypertext thinking).

Scanning technologies now enable us to upload text directly from books* and translated into ASCII code where it is manipulable and accessible by cross-referencing. This will pose important questions as to whether the integrity of a book – the information between its covers – should be preserved in favor of throwing everything into the stew, so that a request for information about Shakespeare’s King Lear could leap between sections of two distinct books, one about the King Lear image in all literature, another about Shakespeare’s tragedies, as easily as between chapters of the same book.

Optical disk and CD-ROM technologies already permit us to mix audio, video, and textual material in massive memory capacity. Professors could videotape lectures for replay, although the very concept of lectures themselves is likely to alter. Libraries of movies, tv science documentaries and interviews, such archived material as the Feynmann lectures, might all be uploaded and with some sweat, integrated into the system. Students would find that the need to learn computing languages or systems of linear commands (as MTS requires) in order to access archives or communicate via e-mail would disappear in favor of visually organized intuitive interfaces that require no special skills.

The main point is, though, that the system is expandable; it is tied to a concept deeper than individual systems of hardware and a concept of education more durable than the limits of educational software, which has been slow in coming anyway, as a recent Business Week (Oct, 1988) pointed out. I think because the investment of hours of labor per minute of presentation time is simply not worth it in most instances, let alone the lack of institutional support and reward for such non-research activities.

Hypertext systems, by contrast, require a fundamental reorganization of how one organizes and accesses information. And though to get a truly sophisticated archive like the one I describe above would require massive commitments of labor, a more primitive version, lacking some of the fancier features of stop-video accessing and commentary, could conceivably be implemented today.

But I would argue that it would be more efficient, more effective, more persuasive, more honest, more totalizing a re-designing of the university’s learning environment to plan to implement such a system for ten years and then make the leap all at once, say in 1999. Imagine if Rensselaer was the first on-stream hyperversity for the third millenium!

Hyperstudents, hyperprofs, hyperadministrators, and the Institute Requirements

The new Hyperstudent: In the hyperuniversity students are led to take an active and responsible role in choosing their course of study. They must have an inquiring mind in order to make the hyperlibrary work as anything more interesting than a dictionary. They are constantly in dialogue with their hyperprofs as they struggle through various pathways in the hyperarchives; they may often need to check with profs as they begin to explore solutions to the problems profs have set them. It is conceivable that their education will be come to be viewed as one long intertwined thread of voyaging through the archives (exploring, for instance, a new application for superconductivity, or a model for an artificial intelligence, or the ideal wing of a space vehicle, or the structure of a new polymer, or a new sintering procedure, or a new design for a bridge, or the structure of sub-atomic reality, or the writing of a software procedure for simulating a game, or the simulated marketing of a new product, or the organization of a hypothetical corporation, or the design of a new computer interface, or …..) rather than a series of incremental little steps, with their sweaty and trivial little hurdles and obstacles, concern about grade point averages, and next Monday’s tests. Though certainly the student will have to meet deadlines, even deadlines can be viewed as milestones. More and more, the hyperstudent will find his or her professor the source of information not found in any text or archived material, a collaborator in the building of a set of answers to be framed to specific problems.

The new hyperprof  finds that his or her more idealistic pedagogical functions and performances have been stripped of some of the more obnoxious tasks: he or she is no longer a gatekeeper, clock watcher, or administrator of obstacles. Rather, as implied above, he or she is now a combination expert resource, human reality checker, problem designer and setter, collaborator, and courseware designer or knowledge engineer, where designing a course may mean nothing more than posing some interesting problems – solved or unsolved – to students based on their level of skill or expertise or curiosity. The professor also becomes the standards controller (or quality control expert) who designates a system whereby the student can measure his or her relative success, if such a measure is deemed possible or even desirable (as it is in most circumstances).

The hyperprof will often meet with students but in varying venues and with more various goals. Large plenary sessions, lectures, presentations, etc. will be necessary as context-settings for different courses and for creating a socializing experience for students. Smaller discussion groups and seminars will be important for discussing special problems, designing teams for problem solving, and more intimate instruction. For many kinds of instruction, nothing will ever replace the classroom environment: for Socratic dialogue, for the creating of feeling and sympathy, for the establishment of shared values, for demonstrating styles of thinking and researching and creating, for performing and demonstrating certain procedures, for sharing experiences, for creating a spirit of competition and collaboration. The one-on-one meeting will often be crucial, especially as students find their way into uncharted territory in their search for answers to specific problems. And, of course, students can meet professors through the computer: through e-mail, through videotapes, through leaving a marked trail of their course through the computer which the professor can monitor (if parties agree or the prof makes it part of the course requirement).

Almost 50% of the new hyperadministrators might find they are, in one form or another, archivists or librarians, although the definition of the hyperlibrarian will be expanded to include software and hardware experts, artificial intelligence designers, communications experts, videotape artists and technicians, information guides, courseware designers, teaching assistants, research assistants, etc. Other administrative functions ought to remain the same, with the need for getting, maintaining, administering, apportioning and adding resources, fiscal, physical, and material, attracting students, sponsoring programs, communicating with the community at large, alumni, parents, new organizations, etc., arranging the calendar, scheduling meetings, etc.

Finally, we come to the shape of the student’s four-year career here, best expressed now in the bureaucratic mess we call the Institute Requirements. Since many of the disciplinary boundaries will be dismantled by the hyperversity system, there will be no need to require a distribution of courses in one or another domain. Students will find quite naturally that to find out everything they need to know to solve a problem or to research a subject or phenomenon, they will ultimately confront questions in domains represented by most, of not all, of what we now call the five schools. Similarly, the notion of a normal progress of courses from simple to complex will either be deeply imbedded in the hypertextual route a student follows in pursuing a topic, or else a trivial and irrelevant question.

Either a student already knows what he or she needs to know, or they will soon find out, or they will fail in their mission (because they couldn’t understand, didn’t ask the right questions or didn’t care to understand or ask). The idea of declaring a major will still hold some value, al though the idea of a ‘cum’ may not. In the place of a transcript of courses, a student could simply request a printout of the map of his/her tour through the archives, marked by dates and times, along with professor’s comments and evaluations appended at important landmarks along the route. The registrar’s function as gradekeeper ought to be obsolescent in the hyperversity, subsumed into the more enlightening task of tracking a student’s progress. A list of problems solved, devices designed and/or built, questions answered satisfactorily or not satisfactorily might also be supplied.

University as mini-universe

In the end, this is a vision of the university as a mini-universe we invite our students to explore with us.  Technology and the new way it invites us to interact with knowledge and learning turns the university as an open, interconnected, ever-expanding book – or multi-media event – rather than a forced march down a few canals or channels. I believe it is a much more enlivening, emboldening, and enticing view of how we can fulfill our mission to teach, learn, research and treat each other as humans into the next millenium.