Association of Research Libraries (ARL®)

David A. Hoekema

Academic Dean and Professor of Philosophy Calvin College

(Or, If You Lie down in the Middle of the Data Highway, the Search Engines Passing By Will Still Generate Multiple Hits)

Philosophers tend to think in abstractions. Let me begin by flouting the stereotype and telling a few stories.

The first was relayed by the originator of the Intelex "Past Masters" electronic texts of major philosophical works. In the course of several exchanges by e-mail in recent weeks, I asked him for examples of the ways in which philosophers' work had been affected by the texts that he publishes on CD-ROM and disk. In reply he cited an American Philosophical Association convention at which a senior scholar -- a recent President of the Eastern Division of the APA--dashed out of a colloquium to Intelex's display in the book exhibit area and asked the representative to run a quick search for a certain phrase in Locke's Essay Concerning Human Understanding. The search took just a minute or two: no, the phrase does not appear anywhere in that sprawling text. Back to the session went the inquirer. An hour later, a somewhat dejected philosopher came to the booth and requested the very same search. At his session, he related, he had cited that phrase as occurring in the Essay, and used it as a basis for his interpretive comments. But at the end of his paper he had been ambushed by a senior colleague who claimed confidently that the reference was mistaken.

The second episode I will retell occurred just last week as I was preparing this paper. I read in the Chronicle of Higher Education of a new search utility at the Library of Congress, and decided to see whether I could give it a test drive using my Gopher server. As a Gopher neophyte I did not know where to look for LC resources, but I found a young woman named Veronica who led me quickly to a menu where I found the option I was seeking. Out of curiosity, I was timing my search: within 4 minutes of accessing Gopher, I was at the very door of the LC's newly opened electronic hall of marvels. Alas, I had no key. "Missing parameter--server terminated--core meltdown imminent," mumbled the doorkeeper, or words to that effect.

On the same menu I found another promising option: information on an LC conference on Project Gutenberg and electronic texts. I asked to have a look. Five minutes passed, then ten. Finally the document appeared. It looked promising, but I was due home for dinner, so I mailed myself a copy, exited, waited a few minutes for the mailman to arrive, extracted the file to disk, and asked a DOS utility to print a copy for me. Again, a long wait. I killed the print job, which brought the printer miraculously to life and produced the first ten pages. Meanwhile I began copying the document to a floppy to read at home. No such luck: it was too large for a floppy disk. What I thought to be a brief summary of a conference turned out to be a half-megabyte monstrosity, although nothing on the menu or the means of access gave me any hint of its size. Rather as if I had found a slim monograph with a promising title on the library shelf, only to discover, as I tried to slip it into my backpack to bicycle home, that it was really a 12-volume work weighing 60 pounds.

Electronic access, I conclude, is both more useful and less useful than it ought to be. We have a long way to go before it fulfills its potential as an aid to scholarship, but it is also serving unforeseen purposes. Electronic aids intended to assist careful study prove useful for professional sniping. New opportunities for communication also serve the needs of data thieves and plagiarists. Remarkably user-friendly and flexible access structures such as Gopher are prone to failure and can yield booby traps.

But I could balance every story of misuse or frustration with another story of how electronic links have speeded and enhanced scholarship. Here, too, I can be personal. Two weeks ago I finished an essay on ethics in teaching, and within half an hour after finishing my revisions I had not only submitted the manuscript to the two editors, at different universities, but had received an acknowledgement from each that it had arrived safely. More impressive still, a colleague in mathematics told me last week that he had just received page proofs from a research journal published in Germany, to which he had submitted his 100-page manuscript, studded with mathematical formulae, just a few weeks earlier via e-mail. The only inaccuracies he found turned out to be his own mistakes.

What is the state of electronic publishing and information access on campus? To judge by my medium-sized Midwestern liberal-arts college campus: it is a remarkable hodgepodge of sophistication alongside ignorance, of highly developed and widely used systems alongside the most primitive beginnings. And the places where each of these can be found defy all stereotypes. Members of one of our natural science departments are finally scheduled to get personal computer's this year: they hadn't asked for them before, and never saw much need. (They already had computerized laboratory instruments that cost more than a lifetime subscription to every philosophy journal published.) But one historian was combing the campus last year in search of a fast enough 486 to run the CAD software that he needs to use to study artifacts from an archaeological dig. (The Engineering Department agreed to let him borrow one.) One member of the support staff in my office is so adept with computer programs -- including an administrative mainframe system that seems to me to have had "user intimidation" as a principal design criterion -- that she is regularly called on to tutor her bosses. Another support staff member in another office, I learned last week, has been using WordPerfect for six months but hasn't yet quite grasped the concept of filenames. Half a year's work is saved in one gigantic file.

But enough of stories. Let me turn to the questions that the organizers of the conference asked me to address. What are the principal needs of scholars for information that can be delivered electronically? What is it that scholars and researchers need, and electronic publishing channels can provide?

To answer the question, let me suggest that we divide the information used in scholarship into four categories: first, bibliographic data; second, quantitative data; third, ordinary text; and fourth, multi-media material. The situation on campus is quite different in each of these categories.

First, consider the computerization of bibliographic material, including both library catalogs and indices of publications. Here the computer has already won the contest with printed information. Go to any of the dwindling number of libraries where the computerized catalog duplicates the card catalog, and count the number of people using each. Computerized access is quicker, more convenient, and in no way less useful than flipping through drawers or paging through bound volumes. Scholars use printed bibliographies when they are not available online or on disk, but not when they have a choice. This change appears irreversible, especially since -- once the initial transition is over, and all the records transferred -- electronic catalogs and bibliographies are far easier to maintain, correct, and update than are their printed equivalents.

In this area, and this one alone, there are few important unmet research needs that are not in the process of being addressed. Some libraries maintain computerized records only of new acquisitions -- a hybrid system in some ways less satisfactory than the old one. Foreign language materials need more consistent standards and more convenient access. Some bibliographic sources are priced so dear that they are difficult for small libraries to afford -- the major bibliographic tool in my discipline of philosophy being a noteworthy example. But these problems have obvious solutions. The challenge for the future lies principally in melding and integrating various information sources. A researcher of the next generation may be able to search the ten nearest university libraries, or six different bibliographic databases, with a single natural-language command, rather than query each one separately using a primitive Boolean syntax.

What of the second area of information: that of quantitative data? Here I have in mind census data, survey results, experimental data in the natural sciences, and the like. Computerized analysis of quantitative data was one of the earliest applications of computers. One of the benefits of increasing speed and memory of microcomputers has been the transfer of statistical analysis from mainframes to PC's. Such analysis has proven to be essential to both the natural and the social sciences. Electronic exchange of data is not yet as familiar or as routine as the exchange of texts, but its importance to scholarship is increasing. If I pass over this application quickly it is because of my limited experience and knowledge, not because I judge it of secondary importance. Those who work regularly with quantitative data will need effective means for their dissemination and consistent standards to enhance their usefulness. Whether there are significant technical hurdles that need to be negotiated in meeting these needs is a question that I am not able to answer.

Third, and closest to the heart of research in the humanities and the theoretical side of the sciences, is the category of ordinary text -- from research notes to encyclopedias, from introductory textbooks to the most abstruse articles. Such texts are the lifeblood of the university, the source as well as the product of the faculty's wisdom. In what ways has computerization changed the traffic in texts? In this case I will subdivide my category into three subcategories.

First, consider texts that circulate actively for a short period after their creation: reports of research in progress, drafts of articles, comments on recent work, reports on recent developments. For such texts, electronic exchange offers enormous advantages over the alternatives, and it is already very widely employed. The Internet, like the cash machine, is a convenience that we lived happily without until we learned to use it -- but today we could not imagine life without it. The rapid growth of such exchange has created unforeseen problems of privacy, accessibility, and quality control, and it has challenged the very notion of copyright in a published work. Not every scholar yet has access to electronic means of communication. At the small Midwestern college where my sister is a faculty member, there is one terminal in the library with access to Internet, and faculty members line up to use it. With each passing month, the number of e-mail users in the academy grows by a large margin, and those still outside the net feel more isolated.

But in a second category of texts--the fruits of research in larger pieces, in more final form -- the reign of the printed word has not been seriously challenged. Scholars prefer to read journals and monographs on paper, and electronic communication has been most successful when it simply serves as a means of producing printed text on demand. There are electronic journals in various fields, it is true, but they remain largely the domain of hobbyists. Sony's digital book, introduced with great fanfare, has been ignored by a public that insists stubbornly on buying words between bound covers. Scholars are if anything even more traditional when it comes to studying their colleagues' research work.

Yet we can mark off from these works of current scholarship a third category of texts in which electronic dissemination has gained a large base of users already: the primary sources of literary, historical, and philosophical study. Ironically, the greatest success of electronic text distribution has probably been in the most ancient texts: the Thesaurus Linguae Graecae is an indispensable resource for classicists, and the Latin corpera now under development have already attracted wide attention. I have already mentioned the philosophical texts that can be purchased on floppy disk or CD-ROM: they include the British philosophers from Hobbes to Mill, translated writings of the Continental Rationalists, and the works of Aquinas, Descartes, and Kierkegaard both in the original and in translation. The list of literary works available in searchable electronic form grows longer each month.

Why are scholars willing to tolerate the eyestrain of studying Hobbes or Jane Austen on screen, yet insistent on bound journals, or at the very least printed copies of electronically distributed journals? The reason lies simply in differing patterns of use. Any text for which one might have reason to use a concordance is a text highly suitable for electronic distribution. Computerized search procedures amount to an infinitely flexible concordance, written to order for each user. The principal barrier to wider use of electronic primary texts is probably their cost -- particularly since the electronic texts are almost always used as supplements to, not substitutes for, bound copies of the same texts, open on the desk.

Let me turn to the fourth category: that of multimedia research materials. To the extent that this term is used for "hypertext" informational and instructional resources it is an exaggeration, for most such resources simply rearrange text and dress it up in graphics designed to invite exploration, with images employed as illustrations. I confess to being something of a skeptic concerning the hypertext revolution that some of my colleagues believe we are entering: the value of such applications for instruction is considerable, particularly for those reared on video games, but their use in research tools is very limited. But you may well wish to discount this as an expression of nostalgia more than reasoned judgment, coming from someone who still uses an IBM clone, drives a stick shift, and listen as often to LP's as to CD's. (LP's, for the younger members of the audience, were an aesthetically superior multimedia music delivery system, with a capacity for 17" diagonal high-resolution color graphics on the cover, recently elbowed out of the marketplace by shiny little discs that are easier for the fumble-fingered to use and generate much higher profit margins.)

True multimedia information resources are at this point only in their infancy. Today they include CD-ROM encyclopedias with integrated video and sound clips, interactive CD applications for training and medical diagnosis, and laserdisks linking a feature film to texts, photographs, and recorded interviews to aid in its study. Tomorrow we may be able to "read" a paper in music history on a disk incorporating its musical illustrations, or study an architectural drawing by walking through the proposed building first in daylight and then in the evening. Capacities such as these, within the scope of current computers but not yet well supported by software, will doubtless bring many changes to the scholarly use of non-textual material.

Rather than speculate further on such future possibilities, however, I want to return to my central question: what do scholars want, and need, in an information storage and retrieval system that electronic publication and dissemination can provide? It is important to remember that -- for a profession as conservative as that of scholarship in the humanities, social sciences, and arts -- the first priority of scholars is to find means of doing more effectively exactly what they do now. Hence the importance of the electronic applications that are least innovative technically, such as bibliographies and searchable texts. For philosophers, historians, students of religion, musicologists, and even researchers in the natural and social sciences in much of their work, the first goal of electronic publication should be to make the same tasks that now take an hour in the library feasible in ten minutes at one's desk.

And then -- with the time freed from tedious pursuit of highly specific information in print sources -- scholars even in the most traditional disciplines will discover new working methods and new modes of analysis that can only be undertaken with a computer's aid. Searches and comparisons that would have required a year's careful reading and note-taking may now be completed in a few minutes, then repeated again and again with small revisions as new syntheses and relationships emerge. Traditional methods of interpretation can be enriched by quantitative study. A classicist can confirm her hunch about the tone of an ancient writer's vocabulary by verifying that he is using verb forms common in legal documents but seldom used in drama. As non-text material is added to a variety of texts, new cross-disciplinary insight will offer themselves.

Will the electronic revolution "change everything," as is often loosely forecast? Will the roles of information providers, users, and interpreters change radically on campus? I do not think so. Incremental changes are already evident. The roles of librarians, computer center staff, and faculty members are in a period of fluidity and uncertainty -- with the result that mismanagement in any of these linked areas can have disastrous consequences for all. The computerization of information also offers a new arena in which departments and academic units can compete with each other for resources and for prestige or, if a cooperative spirit prevails instead, can assist each other enormously.

Let me conclude my remarks by identifying one major benefit of the growth of electronic information, then a major cost, and then a major change in patterns of scholarship. The benefit is this: increasing reliance on information accessed and obtained electronically offers a potential solution to the impending catastrophe that seemed bound to result from increasing faculty demands for specialized research materials, on the one side, and spiraling costs of journals and monographs, on the other. By judicious and cost-effective substitution of electronic resources for print materials, both library resources and faculty research opportunities can expand without breaking the budget. Or, to put it differently: in the electronic age, the new university library can be ten times larger on the inside than it is on the outside.

The cost that I have in mind may seem almost too insignificant to mention, but it is one that has troubled me in recent weeks as I evaluated research leave proposals from my colleagues. Increased electronic access to research resources greatly diminishes the incentive to leave one's academic home to conduct research. But the experience of perusing documents unearthed by Gopher is altogether different from that of arranging a visiting appointment and conducting research in another academic community. The Internet is only one of many factors conspiring against the traditional sabbatical away from home. Others include the weak dollar, the two-career marriage, and cutbacks at the institutional and the national level in research and travel support. But the Internet plays a role, too, in making it too easy to stay home and thus to miss the stimulation of new colleagues and surroundings.

And what is the change in scholarly habits? It is a major shift in ways of creatively wasting time. In the old library, you would look along the shelves for the book you wanted, pull out two or three others with intriguing titles, and -- from time to time -- from such distractions would come important new insights. The very proximity of books on the shelf served to suggest relationships and new perspectives. Gopher's shelves are far more easily traversed -- from one end of the world of knowledge to the other in a few keystrokes. Idle wanderings in the electronic stacks may lead to strange and unfamiliar juxtapositions, and seemingly wasted time may yield very different results than when information is arranged on physical shelves.

Some may fear that the electronic information revolution will render obsolete the skilled guides who now aid with information access: the librarians who help users find relevant reference works and resources, the publishers who both select and shape research results, the senior scholars whose work helps others assess emerging lines of investigation. When information is so readily gathered from the corners of the intellectual earth by one person at a screen, library staff may be reduced to maintaining software and aiding novices, rather than lending expert assistance to advanced researchers. Publishers may simply turn from scholarly publication to other endeavors as researchers post their results on networks and refine them through frequent communication.

But this fear is, I think, unfounded -- just as it was unrealistic to expect that installing cash machines would put bank tellers out of work. There are still lines for the tellers, after all -- for the simple reason that not everything we need from a bank can be provided by a machine. Sometimes we need to complete a complicated transaction, or correct a mistake, or simply talk to a human being about how to do something. For similar reasons, in information retrieval, we will always need interpreters, evaluators, and guides, even if our more routine needs will be met more efficiently by typing at a terminal. The revolution in electronic information will be less a revolution than an evolution in roles for the academic community.

Six weeks ago I was in Xi'an, China, and saw there what must surely be the ultimate hard-copy backup system. In the ninth century the emperor decided to make an authoritative copy of the Confucian texts and the Spring and Autumn Annals, a history of China, and had the entire text inscribed on 140 large stone tablets, each of the 560,000 characters verified for accuracy by leading scholars. (Half a mega-ideogram, was the term the emperor's MIS director probably used.) The stone steles still stand in the Confucian temple of the ancient capital, where they were placed in the eleventh century. Storage does not get any more archival than that.

When I studied the classic texts of Western philosophy in graduate school, I had my own printed copy to consult. But the means I would use to search for a passage were not very different from the methods of a scholar in the emperor's court--skimming through the text, consulting my notes, looking for the right passage. Today, if a graduate student has a computer and a few hundred dollars, she can search the whole Platonic corpus in a few minutes, and call up relevant passages from Sophocles and Aristophanes in a few more. Yet even when the whole Western and Eastern cultural patrimony has been mounted in a digital chip incorporated into everyone's pocket cellular telephone, the need for explanation, interpretation, and conversation will remain, and it will not be met by machines. We will still be lining up for the teller windows.