4.4 Toward The Metalibrary
Limitations of Current Information Technology

There are two major categories of information distribution that are of concern to the discussion in this section--the commercial and the scholarly. Word of mouth also conveys information but in smaller quantities and so inefficiently and incompletely that it will have much less effect on the direction of society in future than it did in the past.

In the third civilization, the chief commercial media have been magazines and newspapers in the print category and radio and television in the electronic. These are the products of highly developed institutions, and they are carefully tuned to the desires of their consumers. As with all institutions, their essential mandate is to perpetuate themselves; in common with all commercial enterprises, this goal is best pursued by paying attention to the bottom line. Correctness, completeness, and societal consequences are of little importance to the institutional media unless there is some higher controlling ethical imperative, or unless such concerns coincide with others that have an impact on profits.

Those who control such media have the power to decide whether something is news or not, whether to present factual accounts or editorials, and whether to identify the nature of either to the customers. If they hold profits to be a higher good than, say, truth, then the "news" material their outlets present to the public will be sifted through a profit filter but not necessarily through an honesty filter. In such a context, the trivial--glamour, sex, money, power, violence and self-gratification--are made to appear heroic, for they appeal to the sensual and can sell product. Meanwhile the genuinely heroic--love, kindness, honesty, peace, moral goodness, cooperation, and social duty--become trivialized; because they lack sensual appeal, they lack profitability. Even if the for-profit media do not necessarily set about consciously to change traditional values for reasons of conviction, they do so inevitably for ones of gain.

However, it is the heroic (other-centred) values that forge bonds between people, giving a society meaning and enabling it to be. Sensual pursuits are selfish, individualist, and isolationist--they eat away at the bonds of society, and if unchecked, can destroy it. Yet the survival of civilization is too distant and vague a goal to affect short-term bottom-line thinking.

Commercial television is particularly susceptible to the temptation of becoming the advocate of selfish sensuality because it can present the illusion of being a "hot" or personally involving medium, even though it actually has no feedback mechanism and hence no group dynamics. Its watchers do not actually participate in the events portrayed, but they can be given an illusion they do have the power to do so. Television's goal therefore need be no loftier than the excitement of emotions; it has no inherent mandate to inform, except peripherally.

Originally, it was thought that television's ability to bring the world to one's home would promote global understanding and cooperation, but the medium's individualistic and sensual appeal can have very different results--reduced creativity and scope of world view, and a levelling to mediocrity. Its immediacy brings the random violence of terrorism to the living room, and can even encourage such acts on the part of those who have no hope of any military victory. Consistent with its natural sensual appeal, television has developed an almost continual portrayal (and linkage) of sex and violence. There is no mechanism in place to assess the effect this has on society and on children in particular. At the very least, such portrayal desensitizes viewers to murder, rape, brutality, exhibitionism, and violence. In its sensationalizing of world events, the product of the commercial media can easily become disinformation and not even the people who produce it may be aware of this fact. What is more, objectors have no means of making corrections, because none of the present commercial media are interactive. Ratings tell advertisers how many people watch a program, but reveal neither their reaction to it nor the effects it has on their subsequent thinking and behaviour.

Among other things, television becomes its own reality, overwhelming any message delivered through it. Nowhere is this fact more obvious than in the fate of several televangelist superstars who lost their own message to the glamour and sensuality of the medium in which it was being delivered. Rather than their message changing the world of their viewers, they themselves became television entertainer/stars, fully entering into the life-style and the values of the artificial world in which their performances were being crafted, and their message became emptied of its content as their lives put the lie to what they were saying.

Another aspect of the problem is that the late industrial age media came to be a closed system of entertainment, offering only highly control-belief filtered versions of certain currently fashionable world events. Thus "news" was created and managed as much by the media as by the participants in events. In order to entertain, they focused on the flamboyant, outrageous, and shocking rather than the ordinary, on the negative and dangerous rather than the positive or uplifting. Their nature is to be oriented to conflict and personality, rather than to information.

Some believe that North American news reporting is sufficiently negative to create its own crisis--one of non-confidence in the very society that gives it free reign to operate. These outlets cannot therefore be relied on as factual sources of information. In this view, their product closely resembles historical fiction--a tapestry of fancy hung upon a few threads of fact. Stories constructed within such a medium may reveal more about the thinking of reporters than they do about the real world.

Critics of such attacks will be quick to point out the many benefits of television--its potential for informing, for entertaining, for educating, and for allowing at least vicarious participation in events most people could never attend in person. There are educational channels, family-oriented programs, sports networks, and a thriving public television facility--all these indeed serve to mute the criticism given above. By extending the choices available to consumers, these alternatives to the standard commercial fare also whet the appetite for what television could be if unlimited choice were in the hands of the individual viewer; that is if each person could supply the filters on all available entertainment and news without having everything predigested. That is, television still has the great informing potential it has always had, even though that potential has not yet been realized.

At the same time, and like the other traditional media (newspaper, magazine, and radio), television may also have reached the limits of its particular technology. Available time and channel space conspire together with the profit motive to ensure that individual choices from these media are severely constrained. It would appear that in order to permit unlimited access to basic information, a major technological breakthrough is required--one that allows individuals to control their own information filters.

Turning to another realm entirely, when one considers the available information channels for scholars different problems become evident, for here correctness and completeness are not usually as serious problems as are relevance and information overload. Today's libraries--paper and electronic--are added to by millions of book and journal pages a day. The use of book review digests, cross references, citation indices, and bibliographies can be extremely time consuming and does not always guarantee that the desired information will be found. Conversely, if something is not found, that does not mean the information does not exist.

What is worse, just because something is stated in a book on some library shelf does not guarantee its accuracy. It might be based on poor research or inadequate information. It might be deliberate falsehood, poorly reasoned, obsolete, an opinion the author later withdrew, or such bad scholarship that it isn't even wrong. It is always filtered through some world view before being presented, and this means that academic information is no more value neutral than is television entertainment. Correct information on a given subject may be in such a fragmented and widely scattered form that it is impossible for one person to search enough of it to synthesize an integrated whole from the many parts. Even in electronic research libraries there are several problems that a new paradigm for information storage and retrieval must be able to solve in order to create something that constitutes a significant breakthrough.


The first step toward solving some of the problems with information access is called hypertext, because it adds new dimensions to the referencing of textual material. Its characteristics are:

1. The ability to follow up a book or research paper directly. That is, citations and references are electronically linked to the original work rather than stored in a separate citation index. This means that bibliographies become bidirectional, and their entries have reference threads (citations) that extend forward from the date of publication as well as backward, automatically.

2. The ability for readers of a paper or book attach their own links to individual arguments in a work. These may lead to entirely different threads, if someone else has an appropriate link attached. Naturally, this applies to the original author, who is thereby able to withdraw some points or retract errors, so that they do not proliferate through the literature.

3. The ability to request missing knowledge from others using hypertext. One can ask for threads to be attached directly to the request if someone can either create or find the desired information somewhere in the library. Such requests to fill "holes" can be published on their own or attached to other papers by threads in the same manner as citations or comments.

No such advanced facility yet exists for general use for several reasons. First, no one computer has enough capacity and speed to implement a sufficiently large data base. Second, (therefore) what scholarly data bases do exist are both fragmentary and scattered over millions of computer stores. Third, Internet indexing and searching are still extremely primitive, and there is no ability for users to attach new links to old data either manually or automatically. Fourth, few journals are at this point even available electronically, so the Internet lacks authoritative editorial and peer review processes. This means that scholars have little assistance in determining the reliability and authenticity of electronic data. Fifth, personalized hypertext software on small computers made its first appearance on the market only in 1987 (Apple's HyperCard for the Macintosh), and considerable work is yet required to take such packages from single-user environments through multi-use (in schools and offices) and thence to large-scale operation and general utility status.

However, enough has been done on this concept to make it clear that it is both practical and viable on a small scale. Moreover, the successes of HTML (HyperText Markup Language) and related notations in creating the World Wide Web shows the concept has promise on a large scale as well.

When advanced hypertext does become a comprehensive academic utility, scholars will be able to keep up to date in their fields for the first time in more than a century. Whenever a new work is published, its bibliographic references will automatically generate links to the older works, so that someone viewing an article in the middle of a chain will be able to move in either direction without consulting a separate reference.

Moreover, problems created by vague, incorrect, poorly reasoned, retracted, refuted, or irrelevant papers will also be alleviated. As things now stand, journals employ referees and editors to sift out and reject some of the junk before publication. This has the advantage of screening out most of the truly awful material before it is published. The disadvantage is that it may prevent publication of radically new ideas just because they are new or because the author is perceived unfavourably by the editor. It can also prevent altogether the publication of ideas not considered politically or religiously "correct" by the editorial establishment. The present system also does not allow an author to retract a bad paper once it is in print--at least not in a way that anyone seeing the original would automatically be referred to the retraction.

In a hypertext environment, everything can be published, and it could be left to each user to decide how to sift the material. Screening would be easily done, for an individual account could be set to recognize only the links approved by reputable editors and referees, who do their work after publication rather than before and who are able to alter their approval if they have a subsequent change of heart. This could be done by attaching priority numbers to links that could later be raised or lowered. Editors' links could be removed altogether only if no threads depended on them or if suitable warning were given to persons who had shown an interest by using the link. This would give those users the chance to establish an independent personal link. Naturally an individual's normal settings--to recognize certain editors' approvals and not others--could be overridden in a search for any new links to the field of interest regardless of whether they were on the "recommended" list.

Researchers could create a personal list of whose recommendations or links their accounts would recognize and at what priority. A person could subsequently reject some of these or approve other links if that seemed desirable. These filters would grow and change as individual interests did and would make the total body of scholarly work (including all the junk) look different to each user. This very growth and change in individual filters could also be automated by appropriate software. Such a system could serve research needs at various levels, for editors' links could also have a difficulty index attached. These indices would determine whether particular items would normally be of use to a grade school, undergraduate, or graduate student, or to a professional. People would be free to change their difficulty index on a topical, subject, or global basis as they learn more or as their interests change. Alternately, they could set the system to change it for them according to actual usage patterns.

In addition, fees could be charged both for publishing and for reading material. An author would pay to have work put on the system, but each time a piece is referenced, the author would be credited with a portion of the fee paid by the reader (The rest of the reading fee would maintain the system). This would also filter out some low-quality material, for few people would continue to pay to publish things that no one read.

The hypertext concept is not new; it originated with Vannevar Bush, Franklin D. Roosevelt's wartime science advisor, during the 1940s. It was not feasible to build Bush's "memex," however, and the idea languished until recently. The coiner of the modern term "hypertext," Ted Nelson, calls a hypertext system with this added publishing facility "hypermedia." His long running Xanadu project was an attempt to implement such systems on a marketable basis, but despite having many corporate homes, no commercial product resulted. It seems likely, given the magnitude of the task, that an advanced hypertext system is more likely to grow from the collective efforts of researchers and editors already using the Internet, than it is from the workings of one mind, however fertile and energetic. As it does grow, the challenge will be to maintain the openness of the Internet of the late nineties while still allowing individual scholars to view the parts they need under some structure.

Beyond Hypertext--The Metalibrary

If this were all that the next generation information systems could do, they would be revolutionary enough, for hypertext alone would radically change scholarship, publishing, and current libraries. For instance, paper books and the need to store them could eventually cease to exist. The quality and quantity of information could improve dramatically. It would always be possible to find out if a piece of research has already been done, and every scholar would have access to the most current material. A great deal of time would be saved, both in library searches and in preventing unnecessary duplication.

However, there is no reason to stop with hypertext. A fourth characteristic could be added to information access in order to transform hypertext from a scholars' tool into an everyday appliance. It could be given the potential to overcome some of the problems associated with the commercial news media by applying the scholars' tools to transfer power to create filters to the public at large. That characteristic is:

4. The ability to link all publications, whether commercial or scholarly, and whatever the original medium.

Before moving on from this point, a definition is in order:

An (abstract) metalibrary is the entire collection of a society's data, information, and techniques, together with the means by which it is stored, accessed, and communicated.

The Metalibrary of the fourth civilization is the complete, electronically linked and accessed version of its abstract metalibrary.

Personal data such as what the family has for breakfast, what colour the cat is, and one's preferences in clothes, music and computing languages is not included in a metalibrary, for this definition focuses on the general knowledge and techniques that the culture as a whole uses and communicates. Although every society has a metalibrary of its generally communicable knowledge, it has never before been possible to assemble and index that information in a manner readily accessible to all. Not everyone can physically drop by the Library of Congress, and even those who do might find a search rather daunting (though it, too has a web site, now). From this point on, references in this textbook to the Metalibrary will always be to the electronic version.

Such a library could contain and link textual material (books, articles, papers, and newspapers of all kinds), graphics (pictures, art, posters) and sound (music, radio programs). It could also have integrated forms such as movies, TV programs, recorded concerts, sports events, and daily news, weather and sports from around the world, as well as lessons on every subject at every level in a variety of languages or with universal translating ability.

Some of this has been done or is in process. Java applets and a variety of other browser plug-ins already allow simple sound and video to be a part of Web materials. Little indexing of visual-oriented material has been done thus far, and communications bandwidth would have to be expanded enormously to handle much of it, but what has been done is a move in the direction envisioned here, however primitive it may be. Moreover, the necessary communications capabilities seem to be being driven by a desire for video conferencing in any case, so by the time much indexed video and other live material is available for access, the necessary hardware may well be in place for other reasons.

In the remainder of this book, a hypertext system having this fourth characteristic will be termed The Metalibrary. The difference between the two is that the Metalibrary allows links to all information, not only to text and simple graphics and animations as at present. Moreover, it would serve the general population, not just scholars.

Emerging technology would give the Metalibrary a variety of abilities. Some of the possibilities are detailed below, though not necessarily all would come to pass, for other factors might make them unnecessary or unachievable.

Metalibrary terminals could become voice-activated, allow either large wall screens or book-size wireless portable units, and be capable of displaying text or colour graphics in the same resolution as a printed book. This would make it the preferred publishing medium for such material as National Geographic as well as the Journal of Combinatorics, including all the back issues.

For most purposes, such terminals would have the potential to replace books, magazines, newspapers, television, and the telephone with a single inexpensive appliance. It would be possible to ask one's home Metalibrary terminal, "What was the gross national product of Belize each year from 1972 to 1999?" One might expect the answer by voice with backup hard copy on the house printer--all without getting up from one's living room chair.

A somewhat more "fuzzily" defined request for, say, a comparison of conservative evangelical and Catholic twentieth century commentary on the meaning and application of the first chapter of John's gospel should also be processed to produce appropriate results.

Neither would information have to be confined to a textual form or be statistical in nature. The command "give me the national news, topic government" could result in the wall-sized flat screen delivering a series of news items, editorials, and film clips tailored to the request. Everyone could design their own news, weather, and sports show, with different announcers and different emphases. A hockey fan could have an all-hockey sportscast, and a would-be traveller could see the weather for Hawaii or Nice instead of Des Moines or Bradner.

The chosen announcer need not actually have ever read that day's news before a camera, for sufficient information could be stored on the person's voice, inflection, and appearance for the Metalibrary to synthesize a program with any desired person's image appearing to do the reading. If people want Walter Cronkite doing the evening news on December 12, 2046, they could have him. If they want Marilyn Monroe, her electronic persona could do it instead. One could have a few personal films and voice recordings made and anchor the news for oneself.

User interests as expressed in actual operation would determine to some extent what current items were available, but once the growing technology allowed enough storage, there would be no need ever to remove an item from the Metalibrary once it had been recorded. Someone who had gone fishing could catch up on a whole week's news on returning.

Movies, including ones now shown first in theatres, could be accessed in the same manner. For the usual access fee, "The Sound of Music," "Ben Hur," "Bambi" or "Rocky XXI" could be ordered and shown in one's own home. Parents would be able to instruct their house computer about what, if anything, their children could order. Television shows would be obtained in the same manner, though the lines between the TV and movie industries could become quite blurred. Producers of a given series would advertise their latest creation and the day when each episode would first be available for viewing. Each family could make up its own schedule of movies, news, comedy, drama, hockey or baseball, and so on--watching when convenient for them, not according to any national or local schedule.

Commercials, however, would probably be inserted at viewing time, though a premium might perhaps be paid to bypass this. On the other hand, it might become economical for advertisers to pay viewers to look at their commercials. Ratings would be compiled daily, weekly, and monthly, and would be cumulated on a long term basis on actual rather than estimated use. Such a use of electronically distributed "canned" entertainment might compete for some time with the already ubiquitous videotape rental store, but in the end, the cheaper, more universal, and more convenient of the two would predominate. If the information highway that is the Metalibrary's infrastructure has sufficient lanes (channels) to transmit one or more movies to each home, it would have the advantage of universal selection and easy accessibility.

Books could be printed a page at a time on the screen for conventional reading, or their contents could be acted out by synthesizing characters cast at the request of the user, who could take the starring role personally, if desired. The same is true of school lessons or university lectures that could be studied through a Metalibrary terminal if desired. An interesting task for some future technician would be to make these lectures interactive so that the students' questions would be answered by the synthesized teacher on the screen. More difficult enquiries could be deferred to the next lesson and taken on by a live expert connected to the session. These questions and answers, once recorded, would remain for the next student with similar interests. In the view of some, such a facility could eventually replace schools, colleges, and universities, though implementing this would take longer than would data-base functions. On the one hand, the "teacher" really would know everything available to know; but on the other, there would be no social interaction with other students, and no personal mentoring possible.

Very large screens could be used to download, store, and display (for a fee) great works of art in homes and offices. These would remain until the owner decided to change the pictures on the wall, at which time the rental contract with the owner of the original art would cease (no fixed term). Eventually, an entire house might be decorated in this fashion, with whole walls being massive screens that projected suitable wallpaper and art collections. Three-dimensional projectors would eventually become available and the images of sculptures could also be rented through the Metalibrary. New television shows or movies, as well as live events, would eventually be available in three dimensions; in fact, such technology may well be among the first of these actually used (although it would require a vastly greater bandwidth than conventional movies).

At the same time, Metalibrary services to professionals will be expanded, and the number of jobs depending on banked information will grow. Anyone still having a desk would have a Metalibrary outlet--probably supplied by the same utility as the one at home, but with a smaller screen (or 3D projection volume). For people on the move, a pocket unit would serve as well, but in less space still.

As in other mature industries, the number of information providers (or at least infrastructure providers) would shrink as their scope grows. In all likelihood, three or four competing Metalibrary utilities would emerge to replace the current patchwork of small companies, but customer equipment would necessarily allow reception from each, switching automatically from one to the other as the user requests. To the end user, it would all appear as a single system.

It should be clear at this point that the Metalibrary might prove to be a concept as revolutionary as was Gutenberg's printing press. It could become at once knowledge machine, entertainer, teacher, home decorator, and communications device. While this cluster of functions would develop over time, it is clear that there would be many disruptions in traditional industries, jobs, and patterns of living.

The same utility that is built with the goal of improving access to information--so that individuals can find out what it is that people collectively know--will by virtue of the facilities it offers cause a massive reorientation of several industries and of almost everyone's life. It is uncertain at this point what the effects will be, because the all possible Metalibrary facilities will come into being entirely as described here. However, this examination may provide some indication of the possibilities, given current technologies.

The Metalibrary described here is not just speculation. It could be said to be partially extant already (albeit in primitive form) in the many interconnected networks of government, academic, utility, and industry information systems that even now exist. Furthermore, what have been described here are actually emerging tools and techniques--ones that enable the manipulation of old data and the generation of new information. These tools are not the totality of the information itself, much of which is already available on the Internet via less comprehensive tools.

In this book, the term Metalibrary will normally be used to refer to a metalibrary that has at least some large subset of the tools and facilities discussed in this section. Where there may be some ambiguity, the term full Metalibrary may also be used to emphasize that it is not simply the information content being referred to, but also a set of techniques for universal access.

The Fourth Civilization Table of Contents
Copyright © 1988-2002 by Rick Sutcliffe
Published by Arjay Books division of Arjay Enterprises