Association of Research Libraries (ARL®)

Joseph J. Esposito, President

Encyclopædia Britannica Publishing Group

I would like to begin by thanking Ann Okerson for inviting me to speak here today. Last year I had the privilege of attending another meeting of this group and spoke about some of Britannica's plans for electronic publishing. Part of what I want to talk about today represents something of a status report on some of our projects. I would also like to extend my remarks, if I may, to other aspects of our strategy, in part because we are beginning to see at Britannica how digital media will affect our entire operation: how we develop products, how we distribute them, and the economics of being in the business of publishing intellectually serious works.

When Britannica first began to look into electronic publishing, we started with the obvious assumption: an electronic version of Encyclopædia Britannica should be more or less similar to the print version, that is, it would have the same number of words, and, if we could clear the rights and the technology was compliant, it would have all the pictures as well. Later we began to think of multimedia: we would add audio and video to the product and make the electronic Encyclopædia Britannica in some respects superior to the print version. At this point, someone jumped up and said, let's make it even bigger! So we put together some plans to add new articles to the encyclopedia, since the constraints print puts on length do not apply to digital media.

In retrospect, all of this early planning seems touchingly naive. What we had been assuming was that electronic versions would essentially replace print versions. We assumed that for all the changes in media, our contributors would still write articles the same way, the articles would be edited the same way, and the product would be marketed and distributed the same way. It did not occur to us until much later that our entire business would be destabilized. New media challenges every aspect of the publishing process, from product design to pricing to distribution. We soon realized that creating an electronic encyclopedia was the least of our problems. Coming up with a new business model was the real headache.

Some people were thinking about these things long before we were. As we sit here a great drama is being played out in the pages of the Wall Street Journal. It is the drama of convergence that Nicholas Negroponte predicted would take place. The convergence is of various industries -- computers, telecommunications, and publishing -- which heretofore were thought of as separate and distinct but now are believed to be linked in an age screaming for information. The drama takes the form of mergers and acquisitions. Paramount attempts to merge with Viacom, but the marriage is interrupted by Barry Diller of QVC; but before that battle is completed, Diller's backer, John Malone, sells out to Bell Atlantic. In the meantime, Nynex takes a position with Viacom, the Newhouse family teams up with Diller, and Bell South also dips into its pockets to help Diller steal the bride on her way to the altar. Meanwhile, back at the ranch, U.S. West has invested in Time Warner, Microsoft is in league with General Instruments, and AT&T's Robert Kavner has pronounced that "content is king." The gods walk the earth and mere mortals must stay out of the way or be crushed underfoot.

What is certain is that if the bride is Paramount, the price for her hand is way out of proportion to her charms. In fact, the merger of Telecommunications, Inc. and Bell Atlantic is worth over $30 billion, significantly more than the entire Hollywood film industry. So clearly the bets that are being placed are for more than the right to deliver Eddie Murphy and Arnold Schwarzenegger movies into everybody's home. Nor can the book publishing industry make up the difference: with total revenue of under $20 billion, the entire U.S. book publishing industry is a tiny mole on the backside of the communications industry. No, there is something mighty peculiar going on.

What is going on is a calculated gamble: the bet is that a huge number of information services, most of which are not even invented yet, will come into being to fill the national data highway. After the big boys tie up the rights to Hollywood movies, television reruns, and sporting events, they will come looking for more specialized wares. It appears that we will soon see the creation of the scholar's channel. Yes, our worst fears -- or dreams -- may be realized, and networks, desperate for programming, may soon try to acquire university presses or Encyclopædia Britannica. We may someday find that the world's largest research library is a division of Time Warner.

We can argue whether this is a good thing or a bad thing, but it is certain that, good or bad, all publishers, even academic ones, have to be prepared for this state of affairs. At Britannica, we began to prepare a product that would make Encyclopædia Britannica available online. The name of this product is, predictably, EB Online. It is in alpha testing now and will become available in September 1994.

Our first idea for EB Online was to create a text-only version of the encyclopedia and install it on campus networks. The reason that the initial release would be text-only is that most computers on campus networks are actually dumb terminals, and even the computers that could support a graphics interface generally run in VT 100 emulation mode. So the academic world was not ready for multimedia, and even though things are progressing, that is still largely true today.

Eliminating the illustrations from Encyclopædia Britannica was not something we were eager to do. There are 23,000 illustrations in the encyclopedia, many more than are in most books dedicated to the pictorial arts, and to drop all those images meant diminishing the editorial quality of the product. But even harder to give up than the illustrations are the so-called embedded graphics that cannot be displayed on dumb terminals. Embedded graphics include such things as special characters, musical notation, and mathematical symbols. Ten percent of the encyclopedia's articles would be affected.

What happened next caught me completely by surprise. We did a survey of people working in the academic library community and were told that graphics could wait; the important thing was to make EB available online as soon as possible. Our respondents said that the primary thing was the 90% of the articles that did not have special graphics, and that the academic library world could wait for graphics until the installed hardware base improved.

I was astonished by this market research for a number of reasons. First of all, it was heartening to see that some people still subscribed to the primacy of text and that Mario and His Brothers have not set the standards for academic publishing. Secondly, it was gratifying to learn that there was a strong pent-up demand for having EB made available in electronic form. (It would be an understatement to say that this was gratifying.) But the third point is the most important of all: the level of sophistication concerning electronic publishing in the academic world was far beyond anything we had anticipated. We were very proud of ourselves to have thought to ask.

So we made the decision to roll out EB Online in stages. The first version, 1.0, will be ASCII, or text-only; Version 1.1 will support at least the Microsoft Windows graphics display; 1.2 will support the Apple Macintosh; and 1.3 will support Sun Microsystems' platform X. Our first lesson in electronic publishing was that we could allow our products to evolve with the marketplace and that our customers would work with us to help us evolve.

While we were working on these questions, we were also tackling the problem of getting the encyclopedia onto a university network. We began to work with the University of Chicago, with which Encyclopædia Britannica has a longstanding relationship. And then we ran into a big, big problem: we discovered that to install the database would cost us about ten times more than we could hope to earn in licensing fees. This meant that from an investment point of view, it would take ten years for us to break even. We are all in favor of long-term thinking, but this was simply unacceptable.

It was when we withdrew from the battle to lick our wounds that we discovered Internet. We realized that if we could make Encyclopædia Britannica available via Internet, we would not have to install it onto every interested campus network. Instead, we could develop the product once, and then sell subscriptions to colleges with Internet access.

This is precisely what we are doing. EB Online will run on a WAIS, Inc. server, which supports the Z39.50 protocol. If those technical terms sound like gibberish to you, fear not. All they mean is that we are using a widely accepted standard for Internet access. Our system is up and running now. In January we will set up the first beta site at the University of California at San Diego. We are seeking to identify two or three other beta sites now. The product will be completed by early summer. By using the WAIS software, we will reduce our costs to a fraction of the first plan. These savings will be reflected in our pricing. Currently a print set of Encyclopædia Britannica costs around $1,500, but the online service will only cost about $1 per student subscriber.

Before going any further on EB's plans, I want to return for a moment to the merger of Bell Atlantic and TCI. As I said before, in 1993 anyone who wants EB has to pay $1,500 or more for it. Soon, because of the creation of a national data highway, the cost of information will plummet. Not many people can afford the $1,500 for EB; we reach about 100,000 customers a year. With substantially lower costs, we could make EB available to many more people. Now, I do not want to suggest that universal access to EB will cure all the world's social and political problems, but on the other hand, I do not think it can hurt. For this reason, I do not begrudge John Malone his billions, as long as the distribution infrastructure he is creating is made available to us without prejudice and for a fee that we can afford. If that means that Arnold Schwarzenegger will in effect be helping to make more widespread dissemination of EB possible, than I am all for Arnold Schwarzenegger.

There is a tendency, I believe, for people involved with electronic publishing to think in terms of ultimate solutions. EB Online potentially is one of these ultimate solutions. Ultimately, everyone everywhere will have access to the global telecommunications network. Unfortunately, most people don't live in ultimate places; they live in stopping-points along the way, places like Bayonne, New Jersey and Morton Grove, Illinois. Our experience is local and particular, and it would be unrealistic to expect everyone to have access to EB Online. For that matter, we know that some people who could have access to a network wouldn't want to use it, so it is important for us to develop a strategy to reach these people as well.

This was a startling conclusion for us at Britannica for the simple reason that, for the first time, we understood that in the future -- and that future is 1994 -- there would not be a single edition or version of EB; rather there would be multiple versions, the form of each tailored to customers' needs, wants, and access. EB would no longer be a literal artifact, a set of 32 volumes in handsome bindings; it would now be a database, or a knowledgebase, that exists in some abstract way as a potential product, a product that becomes actual or is given form only when a particular market opportunity comes calling. This is unsettling; it sounds like mysticism or quantum mechanics: suddenly EB doesn't seem so solid.

Although my colleagues are not united on this issue, some members of the Britannica staff are forging ahead to come up with multiple EBs. I will anticipate a question and say that, yes, indeed, there is internal opposition every step of the way. Some members of the Board of Editors are outraged that we are tampering with our patrimony; some members of the direct sales force are annoyed that we are tampering with their paychecks. But we will proceed until we get fired; and that is not so bad, because we are confident that the people who replace us will study the situation and independently come to the same conclusions. At the outbreak of a fire, everyone independently concludes that it is time to get to the exit.

One of the new electronic editions of Encyclopædia Britannica we are creating is called the Britannica Instant Research System. Some of you may already be familiar with this product. One of my colleagues will be demonstrating it here tomorrow morning. The Britannica Instant Research System represents a different implementation of the EB database. It can run on a powerful personal computer or on a local area network; Internet access is not required.

We began to work on the Research System about a year-and-a-half ago. Actually, the project grew out of a scandal. I imagine most of the people here have heard about this. A politically motivated couple in Texas began to fact-check textbooks to undermine their publishers' credibility. No major publisher was spared. Some of the errors this couple found were outrageous. For example, one textbook asserted that the Korean War ended when President Harry Truman dropped the atomic bomb. Another textbook was found to have 500 errors! The Encyclopædia Britannica editors saw that all the errors that were being reported in the press could have been corrected simply by looking things up in Britannica, but the textbook publishers told us that they did not have the time or money to do proper fact-checking.

Thus was born what we originally called the Britannica Fact-checking System and now call the Britannica Instant Research System. The aim of the product is to provide a highly efficient way to look up information in Britannica. We have tested this product extensively. Perhaps the most interesting thing we have discovered is that electronics reduces the time and cost of fact-checking by 70%. We also tested the product against some textbooks. For example, we fact-checked the book I mentioned earlier that had 500 mistakes. We found 700. Just think what your kids are reading when they go to school!

The textbook scandal made us redefine our strategy once again. We realized that we could take the same content -- in this case the text of Encyclopædia Britannica--- and package it in different ways for different users. The Instant Research System is designed for high productivity; it is intended for people who have to look up a fact quickly and then turn to something else, or perhaps look up another fact. The system is not designed for reading long articles; for such an application, nothing beats a book.

With productivity in mind, we made some important decisions about the design of the Britannica Instant Research System. First, we determined that this product should be run off of a hard disk or server and not from a CD ROM drive in order to make the access time as fast as possible. Secondly, and for the same reason, we designed the product for a 486 machine running Microsoft Windows. It should be obvious to people who know their computers that these hardware requirements go far beyond what most people have at home. Which leads us to a third strategic decision. We determined that we wanted to position Encyclopædia Britannica, or at least this particular version of Encyclopædia Britannica, as a tool for professionals. This was a very big change for us, as encyclopedias are traditionally sold as homework helpers for school children. There is something amusing about an 8-year old using the Encyclopædia Britannica to do homework, but, of course, if your children are as bright as mine, you probably made EB available to them before they entered kindergarten.

Suddenly, the solitary Encyclopædia Britannica has three incarnations: the traditional print set, which is sold to families and libraries; EB Online, which is designed for campus networks; and the Britannica Instant Research System, which will be used by businesses for research and to fact-check documents, and by librarians to service their clients' queries. Digital media make the multiplicity of implementations possible. And more versions are in the works. For example, we are now discussing the possibility of building Encyclopædia Britannica into a publishing system, where it could reside with such other worthy titles as the Chicago Manual and Merriam-Webster's International Dictionary; and we have found an entirely new avenue for Encyclopædia Britannica in the world of artificial intelligence, where it is being used to train neural network software. We now view every potential customer as an opportunity to reinvent Encyclopædia Britannica.

But in fact we are not reinventing Britannica; what we are doing is changing the way the information is delivered. The print version has text; EB Online has text: and it is the same text. The encyclopedia may be more useful in one medium than in another for certain things, but it is still the same encyclopedia. I am fortunate enough to have both print and electronic versions accessible to me, and I use both. I use the computer version for look-up and reference; I use the print version when I want to read something at length. Actually, I have three versions: I also have a copy of the 11th edition in my office, which I use solely to impress intellectual snobs who come to visit. I am, of course, being facetious, but we shouldn't forget that books can have an almost talismanic or ritualistic quality: they can be sacred objects, and they serve to sanctify the individual who can claim ownership of them.

We are not reinventing EB, though we are changing its medium; we are not reinventing EB, but we are changing its form; we are not reinventing EB, but we are taking it from the sacred to the profane; we are not reinventing EB, but in some ineluctable fashion we have changed its meaning. At what point do we say that even though we have not reinvented the encyclopedia, we have, through the accumulation of so many formal changes, done something that we may as well call reinvention?

At some point in working with digital media, the media begin to strike back. A CD ROM may at first submit quietly to having the text of a print product poured onto it, but at some point it begins to reshape that text. Suddenly, it is not the same text any more. The medium is not passive; it serves to define its content.

This is an unsettling thought. When you are dealing with an attempt to summarize the world's knowledge, as we are at Encyclopædia Britannica, the implication is that we are somehow changing knowledge itself. We would prefer not to be that ambitious. On the other hand, it may be naive for us to think that an encyclopedia, or any intellectual artifact for that matter, somehow stands outside of time. Encyclopedias and a view of knowledge are, of course, expressions of a particular era, and for the multi-volume print encyclopedia and the world view it implies, it is a bygone era. An electronic encyclopedia is both an accommodation to changing times and an artifact that will serve to change the times.

As we look down the road at future publications, we see the form of an encyclopedia becoming increasingly spatial, and by implication, our notion of knowledge is becoming increasingly spatial as well. We do not scorn chronology or alphabetization; but these ways of ordering events and ideas no longer seem so incontrovertible, so natural. We see the form of an encyclopedia becoming more and more atomistic, more and more suited to electronic search and retrieval; and we wonder what that says about knowledge itself. Formal changes beget substantive changes; in a language whose form includes no transitive verbs, no one can ever kick the bucket. We do not believe we have even begun to scratch the surface of the implications of digital media, but we do know that everything is getting curiouser and curiouser.