Chapter 24—
Cost and Value in Electronic Publishing

James J. O'Donnell

This chapter is perhaps best read through binocular lenses. On the one hand, it is an account of the value and function today and for the foreseeable future of electronic networked texts. But on the other hand, it questions our ability to account for such value and function. In search of the particular, it risks the anecdotal; in defense of value, it expresses skepticism about calculations of cost and price.

I am a student of the works of St. Augustine and shall begin accordingly with confession. For my own scholarship, the single most transforming feature of cyberspace as we inhabit it in 1997 can be found in a warehouse on the edges of downtown Seattle. I mean the nerve center of www.amazon.com. The speed with which a half-conceived interest in a book converts itself to a real book in my mailbox (48 to 72 hours later) has implications, retrospective and prospective, on the finances of my sector of higher education that could well be catastrophic.

If my approach seems whimsical, do not be misled. The real habits of working scholars often fall outside the scope of discussion when new and old forms of publication are considered. I will have some things to say shortly about the concrete results of surveys we have done for the Bryn Mawr Reviews project funded by Mellon, and more of our data appear in the paper by my colleague Richard Hamilton (see chapter 12), but first I want to emphasize a few points by personalizing them.

First, and most important, Amazon books is a perfect hybrid: a cyberspace service that delivers the old technology better and faster than ever before.

Second, my ritual allusion to the paradox of scholars wallowing in information that they do not actually read is not merely humorous: it is a fact of life. The file drawers full of photocopies, read and unread, that every working humanist seems now to possess are a very recent innovation. Photocopying is a service that has declined sharply in price-if measured in real terms-over the past 20 years, and it is certainly the case that graduate and undergraduate students can tell the same

joke on themselves today. Perhaps only full professors today reach the point where they can joke similarly about books, but if so surely we are the leading edge of a wedge.

But abundance is not wealth, for wealth is related to scarcity. This, I think, is the point of our jokes. When each new book, pounced on with delight in a bookstore, was an adventure, and when each scholarly article was either a commitment of time or it was nothing, the mechanical systems of rationing that kept information scarce also kept it valuable. But if we now approach a moment when even quite serious books are abundantly available, then their individual value will surely decline.

I am fond of historical illustration. A student of mine at Penn is now working hard on a dissertation that involves late medieval indulgences-not just the theological practice of handing out remission of punishment but the material media through which that remission was attested. It turns out there were indeed some very carefully produced written indulgences before printing was introduced, but indulgences were among the first printed artifacts ever. The sixteenth century saw a boom in the indulgence business as mass production made the physical testimony easier to distribute and obtain. The "information economy" of indulgences showed a steady rise through several generations. (The price history of indulgences seems still obscure, for reasons my student has not yet been able to fathom; it would be interesting to see if supply and demand had more to do with the availability of the artifact or, rather, was measured by the number of years of purgatorial remission.) But there came a point at which, almost at a stroke, the superabundance of printed indulgences was countered by loud assertions of the worthlessness of the thing now overpriced and oversold. There followed the familiar cycle of business process reengineering in the indulgence business: collapse of market, restructuring, downsizing, and a focusing on core competencies. The indulgence business has never been the same.

A third and last confessional point: as founding coeditor of Bryn Mawr Classical Review (BMCR) since 1990, I may reasonably assert that I have been thinking about and anticipating the benefits of networked electronic communication for scholars for some time now. Yet as I observe my own practices, I must accept that my powers of prognostication have been at best imprecisely focused. Yes, a network connection at my desktop has transformed the way I work, but it has done so less through formal deployment of weighty scholarly resources and more through humbler tools. I will list a few:

1. On-line reference: Though I happened to have owned the compact OED for over 20 years and now, in fact, own a set of the Encyclopedia Britannica, I rarely used the former and rarely remember to look at the latter. But their electronic avatars I consult now daily: "information" sources on myriad topics far more detailed and scholarly than any previously in regular use.

2. On-line productivity information: Under this category I include far better information about weather and travel weather than ever before; access to current airline schedules and other travel information including hotel directories; nationwide telephone directories including yellow pages; on-line newspapers and news feeds.

3. E-mail as productivity tool: The positive impact of e-mail communication on scholarship for me cannot be underestimated. Relatively little of my e-mail has to do with my scholarship, but that proportion is important first of all: news of work in progress, often including copies of papers, and ongoing conversation with specialists elsewhere is a great boon, no question.

4. Formal on-line publishing endeavors: I confess that I use the kinds of resources that Mellon grants support far less than I might have expected. I did indeed point my students to a specific article in a MUSE journal a few months ago, and I browse and snoop, but it was only in writing this paper that I had the excellent idea to bookmark on my browser MUSE's Journal of Early Christian Studies and JSTOR's Speculum -they appear just below the exciting new URL for the New York Times Book Review on-line.

So we, or at least I, live in a world where electronic and print information are already intermarrying regularly, where the traditional content of print culture is declining in value, and where the value of electronic information is not so much in the content as in the interconnectedness and the greater usefulness it possesses. For a work as explicitly devoted as this one is to carrying traditional resources into electronic form, all three of those observations from experience should give pause. In fact, I am going to argue that the intermediacy and incompleteness of the mixed environment we inhabit today is an important and likely durable consideration. Later in this chapter I will return to the implications of this argument. To give them some weight, let me recount and discuss some of our experiences with BMCR. Some familiar tales will be told here, but with, I hope, fresh and renewed point.

When we began BMCR, we wrote around to publishers with classics lists and asked for free books. An engaging number responded affirmatively, considering we had no track record. Oxford Press sent many books, Cambridge Press did not respond: a 50% success rate with the most important British publishers seemed very satisfactory for a start-up. During our first year, we reviewed many OUP books but few if any Cambridge titles. There then appeared, sometime in 1991 or 1992, an OUP Classics catalog with no fewer than two dozen titles appending blurbs from Bryn Mawr Classical Review. (From this we should draw first the lesson that brand names continue to have value: OUP could have chosen to identify its blurbs, as it more commonly does, by author of the review than by title of the journal, but we had chosen our "brand" well.) Approximately two weeks after the OUP catalog appeared, we received unsolicited a first handsome box of books

from Cambridge, and we now have a happy and productive relationship with both publishers. Our distinctive value to publishers is our timeliness: books reviewed in time to blurb them in a catalog while the books are still in their prime selling life, not years later. The practical value to scholars is that information about and discussion of current work moves more rapidly into circulation. (Can a dollar price be placed on such value? I doubt it. I will return later to my belief that one very great difficulty in managing technology transitions affecting research and teaching is that our economic understanding of traditional practices is often too poor and imprecise to furnish a basis for proper analysis. In this particular case, we must cope with the possibility that a short-term advantage will in the long term devalue the information by increasing its speed of movement and decreasing its lifetime of value.)

We began BMCR in part because we already had a circle of collaborators in place. Rick Hamilton had created Bryn Mawr Commentaries in 1980, offering cheap, serviceable, reliable texts of Greek and Latin authors with annotation designed to help real American students of our own time; in a market dominated by reprints of texts for students in the upper forms of British public schools in another century, the series was an immediate hit. It quickly became the most successful textbook series in American classics teaching. I had joined that project in 1984 and in slightly over a decade we had almost 100 titles in print. In the course of that project, Hamilton had assembled a team of younger scholars of proven ability to do good work on a short deadline without exclusive regard for how it would look on a curriculum vitae-textbook-writing is notoriously problematic for tenure committees. This group formed the core of both our editorial board and our reviewing team. If you had asked us in 1990 what we were doing, we would have said that we were getting our friends to review books for us. This statement was true insofar as it meant that we could do a better job more quickly of getting good reviews moving because we had already done the work of building the community on which to draw.

But what surprised us most was that a little more than a year after we began work, we looked at the list of people who had reviewed for us and found that it had grown rapidly beyond the circle of our friends and even the friends of our friends. A book review journal seems unusually well situated to build community in this way because it does not wait for contributions: it solicits them and even offers small compensation-free books-to win people over. If then it can offer timely publication, at least in this field, it is possible to persuade even eminent and computer-hostile contributors to participate. (To be sure, there are no truly computer-hostile contributors left. The most recent review we have published by someone not using at least a word processor is three years old.)

But the fact of networked communication meant that the reviewer base could grow in another way. A large part of our working practice, quite apart from our means of publication, has been facilitated by the Internet. Even if we only printed

and bound our product, what we do would not be possible without the productivity enhancement of e-mail and word processing. We virtually never "typeset" or "keyboard" texts, a great savings at the outset. But we also do a very high proportion of our communication with reviewers by e-mail. Given the difficulties that persist even now of moving formatted files across platforms, we still receive many reviews on floppy disks with accompanying paper copies to ensure accuracy, but that step is only a last one in a process greatly improved by the speed of optical fiber.

Further, in July 1993 our imitation of an old practice led to a fresh transformation of our reviewing population. We began to publish a listing of books received-enough were arriving to make publishing this list seem like a reasonable practice, one we now follow every month. By stroke of simple intuition and good luck, Hamilton had the idea to prepend to that list a request for volunteers to review titles yet unplaced. (I may interpose here that Hamilton and I both felt acutely guilty in the early years every time one or two books were left unplaced for review after several months. Only when we read some time later the musings of a book review editor for a distinguished journal in another field well known for its reviews and found that he was publishing reviews of approximately 5% of the titles that came to his desk did we start to think that our own practice [reviewing, on a conservative estimate, 60 to 70% of titles] was satisfactory.) The request for volunteers drew an unexpected flood of responses. We have now institutionalized that practice to the point that each month's publication of the "books received" list needs to be coordinated for a time when both Hamilton and I are prepared to handle the incoming flood of requests: 30 to 40 a month for a dozen or so stillavailable titles.

But the result of this infusion of talent has been an extraordinary broadening of our talent pool. Though a few reviewers (no more than half a dozen) are household names to our readers as authors of more than a dozen reviews over the seven years of our life, we are delighted to discover that we have published, in the classical review journal alone, 430 different authors from a total of about 1,000 reviews. Our contributors come from several continents: North America, Europe, Africa, Asia, and Australia. By the luck of our having begun with a strategy based in praxis rather than ideology (beginning, that is, with people who had contributed to our textbook series), we have succeeded in creating a conversation that ranges widely across disciplinary and ideological boundaries. The difficulty of establishing working relations with European publishers remains an obstacle that perplexes us: but that difficulty chiefly resides in the old technology of postal delays and the fact that even e-mail does not eradicate the unfamiliarity that inheres when too few opportunities for face-to-face encounter exist.

Our experience with Bryn Mawr Medieval Review has been instructively different. There we began not with a cadre of people and an idea, but merely with an idea. Two senior editors, including myself, recruited a managing editor who tried to do in a vacuum what Hamilton and I had done with the considerably greater re-

sources described above. It never got off the ground. We put together an editorial board consisting of smart people, but people who had no track record of doing good work in a timely way with us: they never really engaged. There was no cadre of prospective reviewers to begin with, and so we built painstakingly slowly. In the circumstances, there was little feedback in the form of good reviews and a buzz of conversation about them, and publication never exceeded a trickle.

We have speculated that some intrinsic differences between "classics" and "medieval studies" as organized fields in this country are relevant here. Classicists tend to self-identify with the profession as a whole and to know and care about materials well beyond their immediate ken. A professor of Greek history can typically tell you in a moment who the leading people in a subfield of Latin literature are and even who some of the rising talent would be. But a medievalist typically selfidentifies with a disciplinary field (like "history") at least as strongly as with "medieval studies," and the historian of Merovingian Gaul neither knows nor cares what is going on in Provençal literature studies. I am disinclined to emphasize such disparities, but they need to be kept in mind for what follows.

After two and a half years of spinning our wheels, with, to be sure, a fair number of reviews, but only a fair number, and with productivity clearly flagging, we made the decision to transfer the review's offices to new management. We were fortunate in gaining agreement from Professor Paul Szarmach of the Medieval Institute of Western Michigan University to give the journal a home and some institutional support. Western Michigan has been the host for a quarter century of the largest "come-all-ye" in medieval studies in the world, the annual Kalamazoo meetings. Suddenly we had planted the journal at the center of a network of selfidentified medievalists. The managing editorship has been taken up by two WMU faculty, Rand Johnson in Classics and Deborah Deliyannis in History, and since they took over the files in spring 1996, the difference has been dramatic. In the last months of 1996, they had the most productive months in the journal's life and on two occasions distributed more reviews in one month than BMCR did. BMCR looks as if it will continue to outproduce BMMR over the next twelve months by an appreciable pace, but the gap is narrowing.

Both BMCR and BMMR stand to gain from our Mellon grant. Features such as a new interface on the World Wide Web, a mechanism for displaying Greek text in Greek font, and enhanced search capabilities will be added to what is still the plain-ASCII text of our archives, which are still, I am either proud or embarrassed to claim, on a gopher server at the University of Virginia Library. Indeed, when we began our conversations with Richard Ekman and Richard Quandt in 1993, one chief feature of our imagined future for BMCR was that we would not only continue to invent the journal of the future, but we would put ourselves in the position of packaging what we had done for distribution to others who might wish to emulate the hardy innovation of an electronic journal. About the time we first spoke those words, Mosaic was born; about the time we received notice of funding from The Mellon Foundation, Netscape sprang to life. Today the "NewJour"

archive based on a list comoderated by myself and Ann Okerson on which we distribute news of new electronic journals suggests that there have been at least 3,500 electronic journals born-some flourishing, some already vanished. Though BMCR is still one of the grandfathers of the genre (Okerson's 1991 pathbreaking directory of e-journals listed 29 titles including BMCR, and that list was near exhaustive), we are scarcely exemplary: it's getting crowded out here.

But meanwhile, a striking thing has happened. Our users have, with astonishing unanimity, not complained about our retro tech appearance. To be sure, we have always had regrets expressed to us about our Greekless appearance and our habit of reducing French to an accentless state otherwise seen in print chiefly in Molly Bloom's final soliloquy in the French translation of Ulysses. But those complaints have not increased. Format, at a moment when the Web is alive with animation, colors, Java scripts, and real audio, turns out to be far less importance than we might have guessed. Meanwhile, to be sure, our usage has to some extent plateaued. During the first heady years, I would send regular messages to my coeditors about the boom in our numbers. That boom has never ended, and I am very pleased to say that we have always seen fewer losses than gains to our subscription lists, but we are leveling out. Although Internet usage statistics continue to seek the stratosphere, we saw a "mere" 14% increase in subscriptions between this time 12 months ago and today. (Our paper subscriptions have always remained very consistent and very flat.) It is my impression that we are part of a larger Internet phenomenon that began in 1996, when the supply of sites began to catch up to demand and everyone's hits-per-site rate began to level off.

But we are still a success, in strikingly traditional ways. Is what we do worth it? How can we measure that? My difficulty in answering such questions is that in precisely the domain of academic life that feels most like home to me, we have always been astonishingly bad at answering such questions. Tony Grafton and Lisa Jardine, in their important book on Renaissance education, From Humanism to the Humanities, make it clear how deeply rooted the cognitive dissonance in our profession is between what we claim and what we do. Any discussion of the productivity of higher education is going to be inflammatory, and any attempt to measure what we do against the standards of contemporary service industries will evoke defenses of a more priestly vision of what we are and what we can be-in the face of economic pressures that defer little, if at all, to priesthoods.

But I will also suggest one additional reason why it is premature to begin measuring too closely what we do. Pioneers are entitled to be fools. Busting sod on the prairie was a disastrous mistake for many, a barely sustainable life for many more (read Wallace Stegner's luminous memoir Wolfwillow for chapter and verse), and an adventure rewarding to few. But it was also a necessary stage toward a productive and, I think we would all agree, valuable economy and culture. I suggest that if we do not know how to count and measure what we do now on the western

frontier with any certainty, we do already know how to fret about it. We know what the issues are, and we know the range of debate.

By contrast, any attempt to measure the value of electronic texts and images or of the communities they facilitate is premature in a hundred ways. We have no common space or ground on which to measure them, for one thing: a thousand or a million experiments are not yet a system. We do not know what survives, what scales, what has value that proves itself to an audience willing to pay to sustain it. We can measure some of the costs, but academic enterprises are appallingly bad at giving fully loaded costs, inasmuch as faculty time, library resources, and the heat that keeps the fingers of the assistant typing HTML from freezing are either unaccounted for or accounted for far more arbitrarily than is the case for, for example, amazon.com. We can measure some of the benefits, but until there is an audience making intelligent choices about electronic texts and their uses, those measures will be equally arbitrary.

Let me put it this way. Was an automobile a cost-effective purchase in 1915? I know just enough of the early history of telegraphy to surmise, but not enough to prove, that the investment in the first generation of poles and wires-Ezra Cornell's great invention-could never possibly have recouped itself to investors. In fact, as with many other new technologies of the nineteenth century, one important stage in development was the great crash of bankruptcies, mergers, and reorganizations that came at the end of the first generation. Western Union, in which Cornell was a principal shareholder, was one economic giant to emerge in that way. A similar crash happened to railroads in the late nineteenth century. Such a reading of history suggests that what we really want to ask is not whether we can afford the benefits of electronic texts but whether and how far we can allow universities and other research institutions to afford the risks of such investment.

For we do not know how to predict successes: there are no "leading economic indicators" in cyberspace to help us hedge and lay our bets. Those of us who have responsibility for large institutional ventures at one level or another find this situation horribly disconcerting, and our temptation over the next months and years is always going to be to ask the tough, green-eyeshade questions, as indeed we must. But at the same time, what we must be working for is an environment in which not every question is pressed to an early answer and in which opportunity and openness are sustained long enough to shape a new space of discourse and community. We are not yet ready for systems thinking about electronic information, for all that we are tempted to it: the pace of change and the shifts of scale are too rapid. The risk is always that we will think we discern the system of the future and so seek to institutionalize it as rapidly as possible, to force a system into existing by closing it off by main force of software, hardware, or text-encoding choices. To do so now, I believe, is a mistake.

For one example: Yahoo and Alta Vista are powerful tools to help organize cyberspace in 1997. But they are heavily dependent on the relative sizes of the spaces they index for the effectiveness of their results: they cannot in present form scale

up. Accordingly, any and all attempts to measure their power and effectiveness are fruitless. For another example: there is as yet no systemic use of information technology in higher education beyond the very pedestrian and pragmatic tools I outlined above. Any attempt to measure one experiment thus falls short of its potential precisely because no such experiment is yet systemic. There is nothing to compare it with, no way to identify the distortions introduced by uniqueness or by the avenues in which the demands of present institutional structures distort an experiment so as to limit its effectiveness.

What we still lack is any kind of economic model for the most effective use of information technology in education and scholarship: that much must be freely granted. The interest and value of the Mellon grants, I would contend, lie in the curiosity with which various of our enterprises push our camel-like noses under one or another tent flap in search of rewarding treats. Until we find those treats, we must, however, be content to recognize that from a distance we all appear as so many back ends of camels showing an uncanny interest in a mysterious tent.