The Electronic. Journal:

A Progress Report*

Murray Turoff and Starr Roxanne Hiltz
The New Jersey Institute of Technology, 323 High St., Newark, NJ 07102

        The concept of the Journal is reexamined in the light of current computer-communications technology. Four forms of electronic Journals on the Electronic Information Exchange System (EIES) are described as divergent examples of potential journal forms that could become prominent In the future. These are (1) an Informal newsletter, (2) an unrefereed “Paper Fair,” (3) a journal which replicates the traditional print-based system, and (4) a highly structured Inquiry-response system. The economic imperatives which will force the replacement of current print-based journals are reviewed. it Is predicted that the next few decades will witness the emergence of networks of small computer systems dedicated to dispersed Journal functions and under the control of the primary communities responsible for the creation of the Information.


        We all think we know what a scientific or professional journal is, and what it means to “publish” a paper. However, new computer-based alternatives to the traditional print-based journal may soon make these concepts outmoded.

      <----------   l  <----------

FIGURE 1. The scientific journal, classical print-based model.

        Figure 1 shows a simplified generalization of the cycle of communications which occurs for print-based scientific journals. The current print-based system for dissemination of research findings is approaching its limits. As the costs of publication go up and the number of articles produced also goes up, two consequences emerge. The number and costs of printed journals and bibliographic services rise to the point where many libraries, let alone individuals, cannot afford to subscribe to many of the periodicals to which they would like to have access. Potentially valuable papers are not published or are severely delayed, in publication by an arbitrary peer review process which must reject the majority of papers submitted to prestigious journals because they cannot afford to publish them [see, for instance, refs. 1—3]. At the same time, readers pay for journals in which the majority of articles are not of interest to them.
During the last decade, information scientists have pointed out that computer networks could be used to establish the “electronic” or “virtual” journal, which would solve these and other problems [ see, in particular, refs. 4 and 5]. Indeed, a current projection of the nature of libraries by the turn of the century predicts:

The rapid decline of the artifact—particularly the printed book—as the primary device for storage and transmittal of recorded knowledge, and the replace.’ merit of these artifacts with DATA…accessible electronically. [6]

        We are still at least a decade away from substantially supplementing print-based journals with electronic ones, but the first electronic journals are here. Currently in experimental prototype forms, they are a byproduct of the “Operational Trials of Electronic Information Exchange” sponsored by the Division of Information Science and Technology of the National Science Foundation.

        In contemplating the application of computer technology to improving the timeliness and efficiency of dissemination of scientific information, however, it is important that we recognize that the scientific journal is not merely a way of disseminating information, but is also acentral element in the social control system of science (see Fig. 2). The peer review process serves to filter and validate scientific findings. By choosing to publish only what is considered the “best” scientific work, readers are spared the task of trying to select and judge for themselves what are the most important contributions among an unreadable number of submissions. Second, it serves as one of the central mechanisms in the allocation of scientific prestige and of resources for scientific research. Having one’s work published in a prestigious journal tends to increase the resources subsequently available to a scientist for conducting further research. As we shall see in the case history of the “mental workload” journal, any replacement for a traditional print-based journal must take into account that the journal is a social control system as well as an information dissemination system.

        The EIES system, which serves as the host for these journals, is described in detail elsewhere. It provides messages (private communications), conferences (group or public communications), notebooks (an individual composition and file space), and a variety of advanced features. Our purpose here is to describe some of the features which have been developed during the last year to facilitate electronic journals, and the forms which these journals have taken. We will also touch upon some of the human and social factors which will be important to the future success and acceptance of electronic journals.

Technological Features for an Electronic Journal: Submission, “Reading,” and Public Access

FIGURE 2. The classical journal as a social system.

        The SUBMIT and READ functions which are used work as follows. Anyone on EIES who has written a paper can execute a SUBMIT command and identify the locations of the abstract and pages of the paper. The resulting text item that the submission command creates may be used as a message to individuals or placed in a conference or notebook. Anyone having that text item printed out as part of their normal communication process will be presented the abstract for the paper. Then the receiver may execute a READ command referring to that text item and the, whole paper will be printed. This capability lets ‘an author opens his or her notebook to a selected audience.

        A person READing a paper may also choose to do a RESUBMIT command which allows him to give access to the paper to other individuals. The author would not know who has been given the paper by this RESUBMIT function. This means an “editor” can pass the paper to reviewers without the author knowing the identity of the reviewers. However, the author can modify the master copy in its original location and the modified copy is what any reader will see who uses the READ command after the modifications occur. The author may also delete the original item created by the SUBMIT function, and this will also close off access under the RESUBMIT function.

        Another component which is necessary for an electronic journal is “public” access or dissemination, beyond the limited membership of the EIES system. This is provided for by recently developed “public membership slots,” which will allow up to 1000 individuals to share one membership slot. Shared memberships will only allow a user to access defined public conferences and public notebooks comprising the particular journal they are authorized to obtain, They will not have access to the remainder of LIES. The computer will, however, remember what they have .and have not seen and keep them updated on new material automatically. One such public electronic journal is described below. By combining a public conference (where anyone can read or write) and a public notebook (where the writers are restricted to a defined group hut any9ne can read), one can create a journal that is an appropriate mix of contributed reviewed items and spontaneous reader reactions.

        A number of features have been created on EIES which facilitate the following processes:

(1) An author writes and revises a paper in electronic form, using text-editing features.
(2) When it is finished, he submits it to an editor.
(3) The editor reads it and then sends it to the review-ers in a form in which the name of the original author is not indicated.
(4) The reviewers, author, and editor may discuss revisions.
(5) When a paper has been revised and accepted for publication, the abstract is placed in an electronic journal or “Bulletin,” where it is potentially available to thousands of readers.
The “Electronic Journal”

        When EIES was first designed, there was a single mode proposed for all “electronic journals.” Three years of operation and’ hundreds of thousands of pages of text later, we realize how wrong we were. Our initial thoughts were very much along the lines of mimicking a formal journal and imposing this structure on all’ "Bulletins" or journals on the system. What has evolved, however, is a multitude of alternate subfunctions from which user groups can piece together the type of "Journal" operation that satis-fies their needs, desires, and norms.

        The SUBMIT and READ command provide the building blocks for the emergence of electronic journals on the EIES system. There are currently four prototypes in existence or in the development stage.


        The simplest journal is Chimo, a newsletter with items about the members and groups on EIES and new system features. It uses the READ feature for its “supplements”: fulllength papers that have been keyed into EIES by its members, which are then made available to members.

        Chimo is usually a two-page weekly publication. However, many of the short items in Chirno make reference to longer items which the user can select from public conferences or public notebooks by using the READ feature. In a very real sense, it is a current awareness abstract and headline service. Chimo utilizes an advisory board whose members act much like reporters and an editor who has final authority on the actual material published each week. It can be compared with the special interest group newsletters of many professional societies. Chimo does have the advantage of maintaining for each of its readers a marker on the last issue read, so that a reader can get all the unread issues in one operation.  All the past issues of Chimo are stored online and are available for retrieval by standard text searches. The same public notebook that stores past issues of Chimo also stores the more de-tailed items referenced within the Chimo issues.

Unrefereed Free for All

        There is also a public conference called “Paper Fair” which can be considered a totally unrefereed journal. Any member of the system can put a paper into Paper Fair, and any member can READ the papers there and enter their reactions or comments into the Paper Fair conference. Similarly, there is an unrefereed “Poetry Corner” which can be contributed to, read, and commented on by any member of EIES. Still another example of unrefereed electronic publishing is provided by an author who is submitting chapters of his fulllength book within the conference of his scientific community, which the other participants may choose to print out and comment on.

        The role of the editor in Paper Fair is to guide authors wishing to make submissions, to index the submissions in the table of contents for all entries, and to exercise deletion capabilities if ever called for. The typical Paper Fair submission is a draft prepared for a professional meeting or presentation and/or for a journal. Usually, these would not appear in print for as long as six months to a year after their submission in Paper Fair.

        The coauthoring of proposals on EIES has been almost as active as the writing of papers and at least one of the proposals written on EIES appeared in the Paper Fair in both its draft and final form. Book reviews have also appeared in Paper Fair. Because Paper Fair is not readily available outside the community of current EIES users, it ‘serves largely for preprint ‘distribution and for soliciting feedback from others. However, it does also reflect a feeling we have gotten from many EIES users of ‘the need for an unrefereed publication vehicle.

Classical Model, with Variations

        The first electronic journal similar in design to the classic print-based journal was “published” in 1980. It is a journal for the research specialty known as “mental workload,” the study of  person-machine interfaces in the operation of complex systems, such as the controls in the pilot’s cockpit or in a nuclear plant. This particular specialty does not now have a print-based journal. The key persons involved in the group’s electronic journal, effort are Tom Sheridan of MIT, the Principal Investigator; Neville Moray, the Editor; and John Senders, the Evaluator who is to assess its effectiveness.

        The original plans for, this electronic journal on mental workload called for it to be, advertised, refereed, edited, copyrighted, and mass distributed—just as are traditional journals. Some demonstration of the feasibility of the concept did occur, but for the most part the plans did not come to fruition. The editor serves to pre-serve the anonymity tradition in journal publishing. When an article is submitted d online (with the + SUBMIT command), the editor assigns reviewers, who have access to the paper without knowing the author. Likewise, the reviewers send their comments to the editor, who removes the identity of the reviewers before sending the comments to the author.

        Whenever an article is in final form and accepted, it can be “published,” rather than waiting for issues at specific times. Another difference (as compared with the traditional journal) is that all comments by readers could be collected and made available to all of the other readers, along with any responses from the author. Such comments on articles can be signed or unsigned.

        The original plans were to advertise the electronic journal in wide circulation print journals, such as Science. Any interested person could subscribe to the journal. This would bring the subscriber the instructions and access code to dial into LIES on a public member-ship slot, with markers being kept on each of the approx-imately 1000 readers who share access to a slot.

        Anyone signing on under a journal subscriber ID would not see the regular EIES interface, but would be welcomed to the Mental Workload journal (or whatever title was chosen). They would be asked if they wanted to read abstracts, search authors or titles, print articles, or comment on articles or the journal system.

        It is not difficult to do this technologically. The results should show a much shorter cycle from completion of research to dissemination of findings. Costs should also be lower, since each reader prints only those articles, of interest. The difficult problem is a motivational one. How do you motivate people to take the risk of expending effort to write and review for an electronic journal which has no established prestige-granting rating in the scientific community? For, besides serving as official archives of research findings, journals also serve as bestowers of prestige.

        As with new print journals, part of the answer is to try to get the initial reviewers and authors to include a high proportion of people who have an established reputation. And, as Roistacher [9, p.23] points out, another “crucial social aspect of a virtual journal is not ‘merely that scholars submit articles, but that they read and cite articles in virtual journals at least as frequently as conventionally published work.”

        It’ is the motivational and political regulatory factors which have prevented publication of this journal. All software, including the public access, was in place by the summer of 1979. However, months went by before a single person submitted an article! Prestigious members of the specialty area were asked to submit articles and many said “yes.” However, the “payoff” to them in prestige apparently was not high enough for them to place writing such articles near the top of their priorities. An additional problem has been International regulations which have prevented the participation of the British members of the group, as will be described below. The journal finally did publish its first two articles early in 1980, but by then the close of its experimental trial period was fast approaching, and the ambitious plans  for a steady stream of published articles, advertising, and public access had to be abandoned.

        A few of the members of the group were not happy with the software [see, for instance, refs. 10 and 11]. Perhaps a more important mechanical problem was that the experiment bad made no provision for typists or OCR equipment, in order be able to accept any manuscript from an author, rather than the author having to key it in himself or herself. Authors are not typists, and should not be expected to learn the advanced text editing necessary to format a finished manuscript. However, the followup questionnaire data from the participants in this effort support the conclusion that motivational factors were paramount. In Table 1, we see that by far the most important reason given by participants in the effort was that “Other professional activities must take higher priority.” Whereas 65% gave this reason, no other criticism or reason was checked by more than 18% of the participants.

        In addition to the internal evidence from the participants in the effort, a comparison with other groups is the strongest in4icator of motivational problems in the attempt to replicate the classical journal electronically. For instance, group 54 members found LIES significantly more “frustrating” and less “productive” than other groups using the same system. And, most tellingly, we have the fact that other forms of’ journals, described in this article, did succeed.

TABLE 1. Number of mental workload group checking reasons as very
important in limiting use of EIES, follow-up questionnaires (N =17). (a)

Other professional activities must take higher priority
11 (65%)
Inconvenient access to a terminal
3 (18%)
The system is too complicated
Trouble with telephone
Trouble with TELENET
I do not like using a computer system like this
Limited night or evening hours
Had some bad experiences
Cost of telephone or TELENET
The conference comments or messages I have received do not seem worth reading
I do not like to type
Inadequate leadership of the group
I am not very interested in the subjects being discussed
Red notebook documentation looked like too much to read
There is no one on this system with whom I wish to communicate a great deal
(a)Source: Follow-up questionnaire, Spring 1979. Question: Please place a check mark or X in the appropriate box to indicate whether each of the following factors has been very important somewhat important. or not important at all in limiting your use of the EIES system.

Tailored and Structured “Journals”

        A subsystem of EIES called Legitech has been operational since January 1978. The design is not like that of traditional journals and has been described in detail elsewhere[12, 13]. Its implications are extremely significant and deserve some further clarification. Legitech restricts itself to three specialized types of text items: inquiries, responses, and briefs. Any member of the Legitech “journal” or network may enter an “inquiry.” asking others to supply information on a specific issue or topic. Any member may enter one or more “responses” to such an inquiry and may indicate whether he or she wishes to receive all future responses to a particular inquiry. As a result, the member of the network acting in a readership role can self-select the information that he or she will obtain in future readings of this journal.

        After an inquiry has accumulated sufficient responses. an individual member undertakes to compile a “brief” of 3—15 pages, which integrates the original inquiry with all its responses. This brief is then published in a public notebook for the benefit of any interested party on EIES and hard copies are made available to people outside of the ELES membership by the individuals involved in coordinating the network operation.

        Legitech represents a highly structured or specialized “journal,” which we feel is indicative of the variations in the concept of the journal made possible by the introduction of electronic technology. Its success centers far more explicitly on the motivation of its core of writers to exchange timely information and share scarce resources than is the case for the traditional scientific journal. The core members are state legislative science advisors (normally one person for each state) who are expected to be able to advise a state legislature on any and all questions involving any field of science or technology. Usually, they have from three to six weeks to obtain information on any particular topic. It would be impossible for them to rely on the mails with this time constraint. In effect, instead of each state having only one professional staff person, the 25 states involved now each have a staff of 25.

Evolutionary Design for Electronic Journals

        EIES, by mixing information and communications into one system, allows the possibility of evolutionary design and operation of an information system. The concept that the form of a journal can evolve over time through continuous redesign in response to suggestions from contributors and readers is possible in an electronic environment. It may be the key to offering the greatest range of opportunities and lead to a host of very innovative and unique electronic journal forms, such as Legitech.

        Perhaps one of the most interesting aspects of the Legitech experience is the demonstration of such an evolution of the users and the structure of the system, including human roles as well as software. After a year of operation, the system was recently redesigned to accommodate many new features the users now feel they want. These are not capabilities they were interested in at the start of the system, when they were unfamiliar with the technology. They have, for example, given up the concept of a “title” for an inquiry and adopted a common set of keywords and phrases that will automatically be incorporated into an index of all items. Inquiries and responses can now be retrieved through the keyword index. In renaming the system “Politechs,” they have also provided the capability of subdividing their 80 or so members into different interest categories.

        Similar to the Legitech operation is the specialized software developed for some electronics standards panels currently on’ LIES. In the TERMS subsystem it is possible to enter a word and have various users supply alternative definitions. Each of these definitions can be voted upon by all the members of the panel. Since these panels require complete consensus before a definition can be put out as a standard, the TERMS system can alleviate much of the bookkeeping, physical labor, and delays that occurred when this process operated through the mails.

        It has been highly’ informative watching EIES users evolve over time toward increasing utilization of evermore sophisticated information tools . Up until now, we have only provided specially tailored indexes for such groups as Legitech. Recently, however, we have introduced a user-oriented version of our internal file and index structure.’ This will allow users to set up personalized indexes where they can point to related items across different conferences and notebooks. In addition, we plan to gradually introduce a concept called “collections,” which will allow groups to cooperate on the formation and maintenance of indexes which can retrieve automatically the text it references., We will also provide facilities for trading “collections” and for notifying authors when an individual or group has collected any of their material. Collections are the starting vehicle or tools for the role of the “research librarian” or “information broker” in what we hope will evolve into an information market-place environment.

Human Factors and the Automation of Existing Traditions

        The initial ideas which people have for use of elec-tronic information exchange technology can often be characterized as automating on the computer the exact communications conventions and concepts that charac-terize the traditional media. Thus, for example, we have “electronic mail” systems which refer to “letters,” “mailboxes,” and even a “postmaster.” When one “mails” a “letter,” it cannot be retrieved for modifica-tion or even deleted before delivery; after all, once a let-ter is dropped into a “mailbox,” there is no way to get it back. Extending this practice to electronic mail serves no useful function that we can think of.
Likewise, there is no reason why an electronic journal needs to have any limitation on the amount of material published, or any fixed publication schedule for new items. Roistacher [9, p.20] suggests that the publication of referee scores can serve the function of assessing “quality” without preventing publication of articles:

The virtual journal’s essential addition to the evaluation process would be that each referee would give an article a numerical score ranging, for instance, from 0 to 100. The referee score would not only allow the virtual journal to publish all papers submitted, but would also allow readers to treat papers as if they were published in a series of journals of differing prestige. Referee scores would be published with the journal’s table of’ contents and could be used as retrieval items in bibliographic information systems. Low scoring articles would tend to be withdrawn until a satisfactory score was obtained.
        Roistacher extends his concept to include the possibil-ity of readers contributing their ratings of articles after reading them. This would allow an author whose “reviews” were poor to be “vindicated” by a wider readership. The scoring of a paper would thus change according to the reactions of the readership and the time interval during which it is in demand. Ultimately, such systems can provide credits to authors based on any suit-able alternative combination of demand and value scoring.

        However, before disregarding all conventions made unnecessary by the new technology, one would do well to ask if there is any definite gain in utility to the users by making such an alteration in their usual habits. There may, indeed be some useful ‘functions served by the prevailing practices. For example, the traditional journal or newspaper appears on a regular publication schedule. And sure enough, our first operational electronic journal, chimo, does too—it is “published” every Monday. There was a discussion of whether it was necessary or useful for an electronic journal to be a “periodical” in this sense.

        Certainly there is no technological need for this, since one does not have to set up a press run or activate a distribution system to disseminate a new issue; discrete items could be disseminated immediately upon acceptance. However, the habits and motivations of the humans in this communication’ system .do seem to support the carryover of this convention. It appears that both the authors and the editorial board need predictable deadlines which provide a motivation for them to schedule a definite time within a week to finish their work (that time, of course, is usually right before the deadline). So, while having publication only once a week’ rather than continuously might seem to slow down the production and dissemination of new Items, human motivational factors appear, ironically, to cause this convention to operate to actually speed it up. In addition,’ at least some readers like the predictability of a new issue every Monday morning, waiting online. They have stated that it has become something of a ritual, like reading the Sunday paper is for many.

        Our generalization is that the design of new systems must take into account the motivations and habits of the humans who create and exchange information, and the reward system of the communities served by a journal, not just the technological’ possibilities.

Technological and Policy Problems

        At the present time, the EIES system could be viewed as “very rich” or as “too complicated,” depending on whether one is an experienced user or a new user. In order to provide public access to an online journal, we have tried to create a minimal set of very simple commands for journal access. However, for potential readers who have never before used a computer terminal or an interactive computer system, even a simple set of access procedures may be’ too confusing to master without a human teacher present at the first session. We still have a lot to learn about how to create a simple but powerful interface and effective written instructions. Another problem is that with “dump” terminals, such as the typical input and output device, the journal cannot include graphics. The difficulty in presenting mathematical notations is also inhibiting. A related problem is that EIES is not currently configured to directly accept material stored on another computer. It has to be keyed into EIES or transmitted through an intelligent terminal or microprocessor. These are technological limitations which can be overcome, but with which the editors, contributors, and readers of the first electronic journals will have to cope for the time being.

        A more serious class of problems is posed by international agreements about data transmissions, and by the cost of access to computer networks, often set artificially high by protectionist mail, telephone, and telegraph interests. For example, the Mental Workload journal initially had about one third of its editorial board located in Great Britain. It was supposed to have a fully binational set of contributors and readers, as well as editors. However, the British Post Office ruled that it would be illegal for British scientists to’ participate in the EIES system through TELENET, because it would violate its agreement with Western Union International forbidding messages to be transmitted over the trans-Atlantic connection. While the difficulty is not clear to any’ of us who have not seen the documents upon which this ruling is based, we suspect that the wording or intent is to protect the profitability of the current TELEX operations against its replacement with computer-based centralized communication systems. The fact that the British scientists might have been provided new and ‘unique opportunities for collaboration with their U.S. counterparts appears to have no influence on the current bureaucratic regulation process, which one supposes was originally established to see that the “public good” would be served. So much’ for the new technology: It is prohibited by laws formulated before it existed, and by vested interests in the old technologies.

        Neville Moray, editor of the Mental Workload journal, considers these policy and cost issues the most serious of all. “Unless an electronic journal is accessible out-side the USA telephone network,” he states, “it is almost an immoral medium; insofar as it generates information of high quality which is systematically unavailable to very large proportions of the scientific community. Some may say, ‘Bad luck to the countries which don’t sort them-selves out,’ but that is not an acceptable solution” [15].

Forces of Change

        In 1965 the number of periodicals in science and tech-nology was estimated to be about 26,000. By 1977 this es-timate had grown to approximately 50,000. The average cost of a year’s subscription had grown from $15 to $30 in the same period . Even a group of libraries sharing resources and budgeting their purchases in this area so as not to overlap would have to spend over p1.5 million a year for a comprehensive coverage of this literature. When coupled with the fact that journal size has also increased and that the rejection rate has gone up consider-ably for the more established journals, one can deduce a number of prognoses for the production and distribution of scientific and technical literature.

        The more central journals have become spread over a larger number of specialty areas In order to remain cornprehensive and general. This results in a smaller probability that a researcher will find more than a few papers in any particular issue that will be relevant to his or her interests. As a result, many commercial publishers have been willing to undertake, highly specialized journals, catering to very select audiences. The commercial publishers appear to have had the flexibility to step into this area more rapidly than the major professional societies.

        A contributor of a reasonably decent technical paper can almost be assured of finding a place to publish if he or she is willing to try a series of journals until a sympa-thetic ‘set of reviewers Is found. As a result, the literature is getting more scattered for the reader.

        The delays between submission and printing of papers decrease the value of the results to the contributors and readership.’ Coupled with the growing cost for the traditional journal service to both contributors and readers, the premises and economic foundations upon which journal publication is based are undermined. The result is that journals are becoming more a mechanism for recording an author’s credits and less of a meaningful device for current information exchange.

        These factors lead to the observation that any eco-nomical and technically feasible’ alternative to the print-based journal has the potential for bringing about major changes in the exchange of scientific and technical information.

        The EIES costs are compatible with estimates by Roistacher [9] for computer-based word processing systems. These figures compare with average traditional journal page costs for composing of $25 per page. For the reader, a $30 investment in a journal would have to pay off with 15 or more relevant articles in a year to be comparable. Some of the specialized journals are costing almost $100 per year.

        This economic trend promises to continue, with the hardcopy journal increasing in cost and the electronic alternatives decreasing at a rapid rate. The logical extreme is the $20,000 minicomputer tied into a digital network and wholly owned by a peer group of scientists as their own cooperative operation within their subfield. Within a decade, such a machine could easily service 200 active daily users with all the centralized data and paper files they might need for regular exchange of information. On such a basis, they would have each made a $100 capital investment and pay about $1.00 per hour of use, including communications.

        Currently, EIES has a number of classes of memberships that are charged different rates [for details, see ref.16]. There is a standard class--  one membership for normal users and other memberships for those using the system in off hours or only a few hours a month as “readers.” For a 5000-word article, we have the following breakdown of cost factors for the complete article:

Inputting at 10 words/mim                            58.75
Inputting at 50 words/mim                            11.74
Inputting from a microcomputer at 300 Bd       1.96
Inputting from a microcomputer at 1200 Bd     0.49

        The latter two cases are also the costs for a class-one member to receive the article on a dumb terminal. This includes TELENET charges of $3. 75/h and assumes an average monthly usage of 20 h on the part of the member. EIES charges a fixed membership fee of $66 per month. Each member is given free storage equivalent to about two such papers and additional ones would run about $5 per month. The “reader” class of user would pay about $2.50 to receive the above paper, but has a much lower monthly fee to belong. EIES is a nonprofit university-based research operation; the fee includes all direct and indirect operational costs, but no amortization of equipment.

        However, EIES currently operates for 700 users on a minicomputer costing less than $120,000. Set up as a journal operation, it could easily support an additional readership population of between 5000 and 10,000 members. In a commercial environment, of course, one might expect charges for such a system to be approximately double that of EIES current costs to allow for profit and marketing.

        By comparison, the current estimate of the cost of a technical report in paper copy is $6.80 [2], well over double the EIES cost. Currently, the cost of composition of a page for, a technical journal is estimated at $25 as a minimum. Our sample paper would then cost about $375 to input to the paper system. No matter how one looks at the costs, the current technology is quite cost competitive with technical journals with readerships in the under 10,000 category. There still remain, however, the problems of introducing this form of publication within the structure of the current technical and scientific information exchange practices. Perhaps one way to overcome this is to consider that within these cost ranges there is sufficient latitude to support ‘royalty payments or credits to authors, based on the size of their readership, and still be competitive with paper-based systems.

        It is, certainly, technically and economically feasible to imagine a host of independent small computer systems started by groups or individuals. They could be tied into a national digital network servicing a multitude of specialized scientific and technical communities, varying in size from 20 to 1000. It remains only to add some intelligent routing to the network or some specialized directory-oriented computers, in order to allow these exchange services to also serve as primary distribution services to others interested in the products of these communities. There are economic motivations, since the creators of the material could impose their own charges to users and bypass most of the current industry with its largely labor intensive overhead costs. The contributors could also be the source for maintaining such things as current summaries of their work, directories of interests, abstracts of their papers, and the papers themselves. In other words, the concept of a decentralized database, where each segment is the responsibility of the creator, becomes a real possibility if adequate distributed processing for information requests can be superimposed on this network of independent host computers.

        Perhaps the most startling inference is that the segment of the industry most vulnerable in the short term to this technology is the secondary information services, such as abstract and current awareness services. The traditional journals have a good deal more status and prestige, which will aid in maintaining their viability, long after they become economically impractical and somewhat ineffective in their current form. In the longer term, however, the whole concept of what a paper is will undergo dramatic change and even the primary sources will be affected. An “electronic paper” maintained by the authors will change over time, based on readership feedback, and only when the paper has Obtained some level of “score” from its peer audience will it be likely to appear in some hardcopy collection. Such collection issues of a “journal” are likely to be highly focused, so that most printed journals will be for libraries, professional users not in the primary subfield represented by the par-ticular collection, and students. Hardcopy journals are going to be less likely to be published on a regular basis and will only publish issues when there are papers to warrant it.

        While it may take a decade to see these effects on the secondary services and two decades to see the impact on the primary services; there ‘is little doubt that dramatic shifts of this nature will occur. One might argue that the average scientist or technical person will not consider joining a cooperative electronic publishing community. However, we have noted a distinct trend among many of the noncomputer types on EIES toward purchasing their own micros. This is not being done for games or because of an interest in computers, but for setting up their own word processing systems. This is, perhaps, the first step along the ‘way to the future we are trying to depict. The important motivations for the end users as contributors and readers are all in place—it really seems to be only a matter of time.


        While Dr. Turoff is primarily responsible for the design of EIES and its evolution, many of the new features such as those reported here are the product of cooperative efforts at designing to meet user needs. Those who have been actively involved in the systems reported here as either designers or implementors are
Chimo: Charlton Price, James Whitescarver, Elaine Kerr, Jack McKen-dree, and Robert Bozilla;
Collections and SUBMIT: Murray Turoff, James Whitescarver, and Robert Bezilla;
Paper Fair: Robert Bezilla and Murray Turoff;
Public Slots: Murray Turoff and James Whitescarver;
Legitech: Peter and Trudy Johnson Lenz, Harry Stevens, Murray Turoff, and James Williams;
Mental Workload journal: Tom Sheridan, Neville Moray, John Senders, Murray Turoff, and James Whitescarver.

        In addition, the unique sort of collective design process’ that takes place on EIES to evolve specialized structures for human groups capitalizes on the extensive set of evaluation and feedback procedures that have been established as part Of the EIES operational environment. While this is the primary responsibility of Dr. Hiltz, the success of the effort is facilitated by the user consultants arid those who take on evaluation roles for the various, groups involved. Finally, the authors wish to thank two anonymous reviewers who made helpful comments on an earlier draft and Marion Whitescarver for her editing assistance.


1. King, D.W., et al. statistical Indicators of Scientific and Techni-cal Communications, 1960-1980, Vol. 2. Rockville, MD: King Research Inc.; 1976.

2. Lancaster, F.W. Toward Paperless Information Systems. New
York: Academic; 1978.

3. Gove, Walter R. “The Review Process and Its Consequences in the Major Sociology Journals.” Contemporary Sociology. 8(6):
799-804; 1979.

4. Bamford, Harold. “A Concept, for Applying Computer Technol-ogy to the Publication of Scientific Journals.” Journal of the Washington Academy of Science. 62: 306-314; 1972.

5. Senders, John. “The Scientific Journal of the Future.” The American Sociologist. 11: 160—164; 1976.

6. Lancaster, F.W.; Drasgow, Laura; Marks, Ellen; “The Changing Face of the Library: A Look at Libraries and Librarians in the Year 2001.” Draft scenario distributed at the conference on “The Role of the Library in an Electronic Society,” University of Illinois, April 1979.

7. Hiltz, Starr Roxanne; Turoff, Murray. The Network Nation: Hu-man Communication Via, Computer, Reading, MA: Advanced Book Program, Addison-Wesley; 1978.

8. Turoff, Murray. “The EIES Experience: Electronic Information Exchange System.” Bulletin of the American Society for Information Science. 4(1): 9-1-0; 1978.

9. Roistacher, R.C. “The Virtual Journal.” Computer Networks. 2: 18-74; 1978.

10. Senders, John W. “The Electronic Journal.” Paper presented at ASLIB/EURIM 4, Brussels, March 1980.

11. Moray, Neville; Stocklosa, Janis. “Electronic Journals: An Editors View.” Paper’ presented at ASLIB/EURIM 4, Brussels, March 1980

12. Johnson-Lenz, Peter; Jóhnson-Lenz, Trudy. “The LEGITECH-EJES Project: Technical Information Exchange and Collective
Knowledge Building Among State Legislative Research Units.” In: The Future of Electronic Communication, AAAS Symposium
Series. Westview Press; 1980.

13. Turoff, Murray; Hiltz, Starr Roxanne. “Electronic Information Exchange and its Impact on Libraries.” Paper presented at the conference on “The Role of the Library in an Electronic Society,” University of Illinois, April 1979; published in the Proceedings of the Conference: The 1979 Clinic on Library Applications of Data Processing, F. Wilfrid Lancaster, Ed. Chicago: University of Illinois; 1980.

14. Hiltz,’ Starr Roxanne; Turoff, Murray. “The Evolution of User Behavior in a Computerized Conferencing System.” Paper presented at the International Communication Association, Aca-pulco, Mexico, May 1980.

15. Moray, Neville. EIES message 5149, May 14, 1979.

16. EIES Access Policy, 1980. Available from Ms. Anita Graziáno, New Jersey Institute of Technology, 323 High St., Newark, NJ 07102.

*The research summarized here has been supported by grants from the National Science Foundation (NSF-DSI-77-21008 and NSF-MCS-78-00519). The opinions expressed are solely those of the authors, and do not necessarily represent those of the National Science Foundation. Portions of this article are adapted with permission from “Electronic Information Exchange and Its Impact on Libraries,” Proceedings of the Sixteenth Annual Clinic on Library Applications of Data Processing, University of Illinois.

Received March 1979; revised 1980; accepted November 30, 1981 © 1982 by John Wiley & Sons, Inc.