Ideas can change the world. Will these? The uncertainties are many. The model that Paul Ginsparg has already brought to life is one that clearly can work, at least under specific conditions. Where a well-defined group of users, all acclimated to the same kind of discourse and even familiar with standard software packages that transmit well by network, concentrate on producing rigorously analytical material, the relatively unobtrusive preprint server can be a powerful tool. Does it scale up? When the group gets larger or more diverse, when research interests start crossing disciplinary boundaries, when fundamental disagreements of method and style are a substantive part of the field itself -- does this kind of communication bog down? Are third parties needed to organize, control, and referee the conversations?
The way to find the answers to these questions is to begin to do, wherever possible, substantial projects of the kind the Ginsparg has outlined. The "hard sciences" are the obvious place to begin, for theirs is the academic culture most dependent on the journal form (to the near exclusion of the monograph) and on timely and wide distribution of resuls of their investigations. But there are areas of the social sciences and even the humanities where similar enterprises can reasonably expect to succeed. What is striking in 1995, five years after the first electronic scholarly journals came on-line, is that there has been so restrained a rush to this new medium. The growth curve has continued upwards, but considerably flatter than that for internet use among academics as a whole. If the model is so enticing, why the delay?
One reason is clearly the economic immunity of the professoriate. The "serials crisis" and the problems publishers have marketing monographs clearly have not yet had an appreciable impact on the authoring public of academe. It is not palpably difficult to "get published": rejection of a given article is taken as a personal, not an institutional, problem, and most articles in which their authors believe find homes eventually. The insulation of scholars and scientists from the economic burdens of their own publications means that they can remain content with the system as it is. A natural tendency to prefer the familiar, usually expressing itself as a concern about whether e-publication will be accepted by promotion and tenure committees, exerts its staying hand as well.
The economic pressure is now being felt not by faculty but by publishers and by librarians. But publishers have no desire to reposition themselves too rapidly into a market which may not be widely accepted and in which they are not sure they can recover costs of production. This leaves the librarians, guardians of the integrity of the system of scholarly communication as a whole. It is in fact in the library community that some of the most venturesome and at the same time pragmatic proposals have come forth. The series of symposia conducted by the Association of Research Libraries since 1992 has given stakeholders on all sides a place to express concerns and inform themselves. The Mellon Foundation published a 1992 study of the state of library budgets (noting inter alia the declining share of overall university budgets that libraries have had the last twenty years) and pointed to electronic publishing as a possible source of amelioration. (See "Hyperlinks" section at the end of this book.)
The Association of American Universities, that is to say the presidents of the leading research institutions, teamed with ARL in 1994 to write strong reports pointing a way forward through judicious use of electronic publication. In the fall of 1994, an ad hoc group of scholars, librarians, and publishers convened in the Columbia University Library to turn ideas in the atmosphere into a concrete series of proposals for ways individual institutions could begin to shift the balance towards economically more viable media. Now in 1995, other glimmers of venture are appearing, e.g., a joint project between Emory University and Scholars Press or another between the Johns Hopkins University and its own Press to put existing journals on-line. The Mellon Foundation is taking a participatory interest in several such projects, to see what the economics of the new medium are really like. It is too soon to say that this will be a groundswell, and the aggregate collection of projects proposed and under way still feels relatively unambitious to longtime observers of the scene. Will it take exacerbated economic crisis to impel further movement?
One of Harnad's economic points is really a sociological point and an important one. The economy of science journal publishing has been constructed so that whatever money changes hands, virtually none ever reaches the authors of the articles. They receive their compensation indirectly, from employers who value their research output. Harnad thus correctly insists that the researcher looking at the published article wants more than anything that it reach as wide an audience as possible, with as few barriers as possible. Thus in this case, traditional publishers cannot defend the need to recover costs from the end user by appealing to their need to reimburse authors. That is a real economic change whose ultimate purport we can now only surmise.
Other considerations will play a part in how this all turns out that did not arise in the course of the focused conversation here. It will not be out of place to suggest a few.
The nature of the scientific or scholarly "paper" or "article" has been shaped by the medium of publication and distribution. Each item must be an independent grain of information, linked if at all by indexes and abstracts and cross-references. There is value (for promotion and tenure) in piling up total number of such grains. We know the difference between an "article" and a "monograph" by several criteria, but a crucial one is length -- if it's too long to be an article, then it should think about becoming a monograph.
But in a world where the artifacts are distributed electronically, numerous changes are somewhere between possible and inevitable. Grains of knowledge will attract each other: if I publish a note this year and add a note next year, at the very least they will be linked to each other dynamically (whether they appear in different "journals" or not) and it may very well make more sense to replace the first note with a longer, coherent version summarizing the two stages of research. All sorts of question of version control and citation authority arise, to be sure, but as reader and as author I think my interests would be served if those can be solved and connections made stronger. Similarly, I need not worry about my article getting too long (in plain text of almost any kind, the difference for costs of storage and transmission between a two page article and a two hundred page monograph will be negligible), and so the article and the monograph may cease to be distinct categories and a new configuration emerge.
Other changes will ensue. If in a WWW document I now need to cite another WWW document, I can make the link dynamic, so that instead of a footnote that says "Smedley, op. cit. 235" I can make a link and let you read Smedley's page 235 if you wish, to see whether what she says corroborates what I say or not. On this principle, not only my own works will be linked to each other, but a growing body of scholarly literature may be enmeshed in a net of links and connections that multiply the value of each item appreciably. At some point when that happens, the intrinsic superiority of the electronic to the paper medium will become ineluctable and the rush to cyberspace will be on in earnest. But we still do not know for sure what it will cost. Debates such as the one included here on the percentage of costs of print publication devoted to printing, binding, and distribution (which almost always omit consideration of the costs to libraries of cataloguing, binding, shelving, circulating, and maintaining) are not well equipped to address a deeper question. How much of what we are now willing to pay for will we continue to be willing to pay for?
Harnad et al. will argue that HTML-tagged texts on the WWW are perfectly satisfactory for reading purposes, but such texts are undoubtedly problematic, not least because the author/editor cannot control completely the look and feel on the screen at the reader's end the way author and editor can now control it. Is that a value worth keeping? It might seem not, but recall how the first camera-ready copy for short run printing in the 1970s, based on photographed typescript, was very reluctantly received, and desktop publishing only really came into its own when the visual quality of the output was able to compete with print. Where is that threshold of acceptance in electronic audiences? It will vary from audience to audience, and for now we don't know.
We also don't know how much technology and support we will need. It is easy enough now to say that the traditional article will take up little or no bandwidth and so place a minimal strain on the network. But what if a scientist begins to believe she needs to give you full motion video, from six different angles, of a crucial experiment? Or if a report on an experiment conducted on the space shuttle feels it must be accompanied by gigabytes and gigabytes of the raw telemetry data sent back to earth? In principle (this is the power of the medium) such full presentations of data are possible and desirable, but in practice, they may begin to add costs (getting that video formatted in a standard way for transmission to all platforms) that we do not now properly anticipate.
For all these concerns, it is impossible not to have at least some guarded optimism about this future. We need not imagine substitution or totalitarian takeover by any of the players on the cyberspace frontier. What needs to be foreseen, instead, is the shift in the mixture of kinds of information available on what terms. There is already a lot of free information in the world (that is, information given away at no direct cost to end user): the phone book for one, all the information in the Borders or Barnes and Noble store that gets consumed by browsers who never buy -- the travel section at Borders in my neighborhood takes a pummelling in late winter as people plan their summer vacations for free, television news (paid for by soap-sellers) for another. Some information, on the other hand (stock tips from really qualified advisors), we pay for at a high price.
But life on the Internet has been marked, at least so far, by a far freer exchange of information by its producers than seems to be the case in the world of print. This "circle of gifts" culture exploits the power of the individual to multiply her words indefinitely at minimal cost of time and effort. So it is that documents which one scholar produces for more or less personal use (say for classroom distribution), or which one hobbyist produces purely for pleasure (say a Civil War hobbyist obsessed with the history of his small town in Tennessee), can now take on a utility they were denied before. Collaborations are possible, like the TOCS-IN project at the University of Toronto where a collective of classicists all over the world each take one or two scholarly journals to track, typing in the table of contents of each new issue as it appears; the whole is greater than the parts, and this easily gopherable resource is now widely used as a way of keeping current in the field, at no net cost to anyone.
It is examples like these that encourage us to dream large dreams about the free flow of information in the future, but even if those dreams fall short of realization, it must be considered likely that the proportion of the total information economy that will be occupied by this free exchange will be larger than is now the case. What this will mean, none can say, but it will influence the market for purchase and sale of information in many ways.
In the end, it is perhaps the word that Steven Harnad started with that begs the question: "esoteric". His idiosyncratic use of the word is meant to highlight publication with relatively few authors and readers, too negligible economically to be of interest to commercial publishers and relatively easy to redirect towards free net distribution. At some level, Harnad is undoubtedly right. Some information will move this way. But "esoteric" as he uses it is a word that defines the position of a certain sector of information as it is produced and consumed in the world of paper publication. When once the power of the networks has been harnessed and a substantial part of the academic population has migrated to using it, a new economy will emerge. What we cannot now predict with accuracy is what will seem "esoteric" there and what will be commercially viable. The lines may very well fall in very different places from where they fall in the world of paper publication.
The responsible course for universities and research institutions concerned about the future is to press the claims of that sector, to experiment responsibly and venture bravely, to see if those lines can be drawn in a way that favors the widest and freest flow of information of a scholarly and scientific nature. In his "Four Quartets," Eliot described the poet's mission as "to purify the dialect of the tribe"; but in a sense science and poetry work to tha tone goal from different directions, seeking by their discourse to enhance the quality of society's common discourse, by rooting out error and imprecision and finding new and true things to say and ways to say them.
To that end, we need a better word than Harnad's, for esoteric suggests something limited and restricted in scope, when what we need now is a conception that encourages us to broaden the category and make it as powerful as possible. The network is a place of dialogue, as this collection makes clear, where the best of what is thought and said enters its own world of permanent transition. It is dialogue in which only a few participate, perhaps, but which is radically open to all who meet its standards of rigor. In that spirit, perhaps we could begin to speak not of an esoteric form of publishing but of a Socratic realm of dialogue.