Vol. VIII, No. 1

   
May, 1995

Publishing with Computers

by H. Eiteljorg, II


Publishing with computers can mean two quite different things. It can mean using computers to aid in the process of creating a standard, printed book or journal, or it can mean making the information available as computer files, with no paper object produced at all. The latter is normally called electronic publication; the former I will call computer-assisted publication.

Computer-Assisted Publication

Using computers to help with the process of making a standard paper publication can save considerable time and trouble, as many who have made their own camera-ready copy from a laser printer will testify. It can also provide more control over the final result, since the number of steps required after the manuscript leaves the author's hands is reduced - as is the need for intervention by trained personnel. By the same token, the absence of experienced and well-trained personnel may yield less than optimum results.

My recent monograph on the older propylon provides a good case in point regarding this use of computer technology.

The manuscript was presented in computer form - as a file on a floppy disk in a word processor format. In addition, all the drawings were created on a computer and available as computer files.

A book designer was engaged by the Archaeological Institute of America to create a design for the monograph that would also be used for future AIA monographs. That design was created with a desktop publishing program.

The printer, however, was not ready to apply computer technology to the physical production process; so the text ultimately had to be set in traditional fashion, and the desktop publishing work could be used only as a guide. Similarly, the drawings were not used in their electronic form but paper drawings were produced on a very high-resolution device and treated as camera-ready art. Unfortunately, the paper drawings were not reproduced well by the printer. As a result, the drawings intended for the book did not have the same density of line as the originals. (I had spent a great deal of time adjusting the lines in the drawings for the proper relative densities.)

The drawings were therefore re-created, this time by making a high-resolution negative so that there need be no photographic reproduction for making the printing plate. This, it was assumed, would guarantee accurate translation of the densities in the images.

Alas, the drawings were still not reproduced accurately. The drawings in the book do not match the originals created from the same computer files on the same high-resolution machine. Apparently, the process of making a print directly yields a certain result, and the process of creating a negative, then a printing plate, and then a print creates something rather different, at least in part because of the differences between ink on paper and the film process used to make the high-resolution originals. Without the involvement of an experienced user of such a process, the problems could not be anticipated, much less solved.

This project shows that using computers to aid in the creation of a printed book may need more care - and more experience - than had been thought. Camera-ready copy demands good reproduction processes; even negatives produced directly from a computer may require some experience if one is to obtain the desired effect. In any case, the ability of the printer to work with computer originals is crucial to the real value of using computers in this process, and, judging by my own experience, anything of real importance should be fully tested before being tried.

Electronic Publishing

Publishing in electronic form is another matter. An electronic publication is not a book or journal one may hold; it is only a series of computer files - whether on one's own computer or on a computer somewhere else altogether. That does not imply less rigorous review and editing procedures. An electronic publication may be as rigorously reviewed and edited as a paper one. The only certain difference is the presentation - on paper in the one version, on screen in the other.

Electronic publication requires care in the presentation, as does making a book, but there is a host of other problems unique to the electronic world.

As one requires assistance to create a book from the text and graphic materials, even if they are in computer form, so scholars will require assistance in creating electronic publications. As a result, the physical appearance of the final product will be determined in part by the skill of the electronic publisher, just as the printed result is determined by the skill of the printer. Furthermore, the screens on which the publication may be seen can differ dramatically from one computer to another, and, given the pace of change in the computer world, one can be sure that the screens of a few years hence will be quite different from the screens of today. Those different screens will create different appearances for text and images. Once the paper book is produced, on the other hand, change is imperceptible (unless the book is printed on acid paper and subject to decay). One may or may not like the appearance of a particular book, but it is stable.

A more worrisome problem with electronic publication is the potential for alteration of the document by users - either innocently or not. Although it may not be possible to change all electronic media, documents can be altered in some cases.

There are two other serious difficulties with electronic publication. First, not everyone has access to the computers needed to take advantage of an electronic publication, and it may be decades before computers become truly ubiquitous. Second, electronic files have short lives; they are not permanent.

Inadequate access to computers may be alleviated by time and by public facilities - terminals in libraries, for instance. But the short-lived nature of computer files is a problem that will persist. This is not simply the problem of magnetic media that lose their magnetism. Although that is an issue, the more critical one is that computer standards change and evolve so that files from the recent past, even if perfectly preserved, become useless. Such files may become obsolete if the underlying computer system changes or if the application software changes. In either case, obsolete computer files are roughly as problematic as samples of languages no longer known. If that is so, and if the files used today may be obsolete in a decade or so, then what permanence can an electronic publication claim? If electronic forms of publication are ephemeral, why would anyone choose to publish electronically?

Given the foregoing, why is electronic publication receiving so much attention today? Are there special virtues, and are there solutions to the serious problems discussed?

I think there are virtues that make this technology very desirable, and there are potential, though not ideal, solutions to the problems. The virtues are both practical and theoretical.

On the practical side, electronic publication is relatively inexpensive. The editorial costs do not change, but the reproduction and distribution costs for an electronic publication are minimal. In fact, some models of our electronic future assume that the material that is published will exist in only one place and not be reproduced at all, simply accessed over a network like the Internet on demand.

Electronic publication is also quick. The entire printing process is eliminated; so is the time required for it.

Electronic publication also makes it possible to eliminate many of the proofing steps, since the comparison of one electronic document to a successor can be done automatically.

There are theoretical advantages as well, and they are more important. We now use many forms of electronic data when we record information. Databases are the most common; CAD models, spreadsheet files, and statistical tables are also widely used. Including the information from data files in a paper publication requires that the data be excerpted, summarized, or otherwise abbreviated. The result is that, at best, a small fraction of what is known is available. In an electronic publication, on the other hand, the complete data files should be a part of the "document;" then the reader has access to the full record, not simply the author's abbreviations. The reader can even perform his/her own analysis if desired. Publishing on paper permits only the analyses of the author, only the views and selections of the author. Access to the author's full data, on the other hand, permits other scholars to ask new questions, to analyze in new ways, and to add to our body of knowledge without duplicating work already done by their predecessors.

Electronic information may also be corrected or augmented relatively easily. As a result, electronic publishing does not create a frozen and immutable item like a book. Errors can be corrected; new information can be appended; even reviews could be referenced. (Changes can be marked to create a record of successive versions of the document.)

With electronic publication there are many ways to provide indexing, links between and among documents or parts of a document, and connections to other electronic material. Those links, particularly those to external documents, make any single document more valuable by connecting it more closely to other sources of information.

If electronic publication has clear advantages, how do we overcome the problems discussed above - unstable appearance, unauthorized alterations to the original, unequal access, and impermanence?

There is no way to overcome the problem of changing appearance on different computers unless we standardize computers. That is unlikely in the extreme. But it is also a modest problem. Some who access a particular document will not see it in its best light, but they will still see it, and they will still be able to get the information - unless the author and publisher have produced a flawed product. That, after all, is not new. Choices concerning quality of production are made now - but only by the publisher and author. The quality of a production in an electronic environment will depend as much on the equipment of the reader as the choices of the author and publisher.

Unauthorized alterations of the original are not a severe problem. (Note the use of the phrase unauthorized alterations. Authorized alterations, as discussed above, are very valuable.) So long as there is an assured original available, any copy can be checked for authenticity. Some distribution systems would make alteration of the original virtually impossible, and encryption could be used if necessary. While someone might alter a file and then pass it on as if it were the original, it is hard to imagine the combination of events that would make such an attempt more than an annoyance.

Unequal access to electronic information is also a problem that will not end but may not be crucial. Libraries will, I believe, provide access through on-site equipment for those who do not have their own computers, and that does not limit those users significantly more than they are limited today by the need to get books from those same libraries.

Impermanence of electronic materials is the biggest problem, and there are no perfect solutions. If published material is ephemeral, . . .

The use of an archive for preserving electronic publication (as the Archaeological Data Archive will preserve data files) is one way to prevent loss of information. In such an archive the files would be migrated as necessary to new forms - indeed, it would often be necessary to maintain information in two or three forms at the same time. That, however, does not help the individual owner of electronic files when the files become obsolete. Nor does it help the library with similar files.

If purchasers buy access to the document in an archive where multiple formats will be retained, rather than a physical copy, that may solve the problem. It assumes, however, that a great deal of material will flow over the networks and very little will exist in individual - or even institutional - digital libraries.

Another solution would be to permit owners of files/documents to obtain new versions any time there is a necessary conversion of file formats. The owner of the documents is then at the mercy of the publisher. If a publisher goes out of business or drops an item from the catalog, how does one update the old documents?

A simple solution would be to reduce documents to printed form so that they could be reconstituted if there were a need to do so. That could be done, even with unusual electronic formats, if the nature of the format were specified. But that process still would leave the purchasers of electronic documents with useless files. (An electronic version of this solution was offered by Jeff Rothenberg in Scientific American, "Ensuring the Longevity of Digital Documents," January, 1995, 42-47.)

Of course, it is also possible to publish in both electronic and standard paper forms so that the purchaser has a choice. In that case, however, the economic advantage of electronic publishing would vanish.

I think the best solution to this dilemma is to make documents available from networked archives. Access over a network could create excess traffic on that network, especially for frequently consulted documents. Also, the policies and viability of the archival organizations would become terribly important. However, the documents would be readily available; they would also be protected from obsolescence, and that is essential.

The conclusion I have reached - that there is no ideal solution to the permanence question - suggests that we take considerable care as we begin to use electronic publication. We must be mindful of the risks, and we must try to provide appropriate safeguards. When we are dealing with material that benefits only marginally from the use of electronic systems, we should think long and hard about whether or not to use electronic systems at all. That is not the conclusion to which I wished to be drawn; the virtues of electronic publication seem to me to be powerful arguments for the technology. So I would have preferred to conclude that electronic publishing is an appropriate technology without serious disadvantages.

Despite my misgivings and those of others who have written about the problem, I think a variety of forces will push us to do more rather than less electronic publishing in the years to come. Care must be exercised.


For other Newsletter articles concerning the use of electronic media in the humanities, consult the Subject index.

Next Article:Archaeological Institute of America Outreach to School Children

Table of Contents for the May, 1995 issue of the CSA Newsletter (Vol. 8, no. 1)

Table of Contents for all CSA Newsletter issues on the Web

Go to CSA Home Page