COMPUTERS and COMPOSITION 10(3), August 1993, pages 49-62

For those of us concerned with teaching college composition, the first decade of the personal computer revolution has been dominated by the wonders of word processing. In the ease with which it facilitates revision, it has been a technology tailor-made to a discipline focused on the writing process. With word processing, it becomes as easy to revise a text as it is to leave it alone; indeed, most of the playful aspects of the computer only come into play when we begin manipulating text we have already entered. While there are other advantages of word processing-- flexibility in page layout; spell-checking and other forms of file-checking; and new, economical, and flexible means of storing and transmitting files--it is this first advantage of editing that has been at the center of the initial attraction to computers for most writers and teachers. Word processing in this sense triumphed completely, among my conservative colleagues who teach literature as well as among more radical teachers of composition, because of its effortless, transparent support for both the theory and practice of writing as process. Or stated slightly differently, word processing, although so effective at transforming practice, did not raise important pedagogical and methodological issues: One could introduce word processing into one's teaching or research and continue working pretty much as always--a condition that explains, with the notable exception of Michael Heim's (1987) Electric Language: A Philosophical Study of Word Processing, the dearth of extended analyses of the larger cultural impact of word processing.

Such had been the case at my own university, where we have been guided from our first efforts in 1986 by the notion that we could help our students the most throughout their four or more years of college by giving all of them minimal training in word processing, specifically in the form of a self-contained two-week instructional unit offered as part of their regular first-year composition class. The first-year composition course would remain essentially unchanged, as would upper division literature classes and other undergraduate courses that require written work, with this one difference: that with access to word processing, all students would be able to produce typed (printed) texts in a technological environment that allowed them to revise (or, in perhaps too many cases, not revise) as they and their teachers saw fit. Word processing, in our scheme of things, was something that students learned from us inside of class--much like the traditional library, research skills we continue to teach in our course--but for the most part did on their own, outside of class. Despite pioneering a computer-based component of our first-year writing class that provided basic word-processing instruction for almost 20,000 students over the last seven years, our use of computers actually did little to undermine the paradigm of the traditional writer, that solitary individual whom Linda Brodkey (1987) describes as "alone in a cold garret working into the small hours of the morning by the thin light of a candle" (p. 396).

This is not to say that we were blind to the larger shift in the profession from an emphasis on the individual and the writing process to an emphasis on the group and collaboration--only that pedagogical changes reflecting this shift were happening in our classrooms largely independent of any computer-based instruction we were offering. We approached word processing, in other words, as a tool for allowing individual students to generate and later revise the texts that they worked on in classes that were increasingly given over to collaborative exchanges. And so things might have continued indefinitely--with our continuing to teach students how to do word processing in a distinct two-week unit largely unconnected to emerging classroom practices--except for the gradual emergence of one new wrinkle in computing: networking.

Although our computer lab had been networked since its inception in 1986, we had been ignoring (except for common printing) any special networking capabilities. That is, we taught as if the computers were not networked, communicating instructions to the class verbally or through handouts and collecting and swapping printed copies of student work (or using low-tech work-arounds such as having students move from terminal to terminal a la musical chairs). In part, we did this for the same reasons one suspects many others do in a similar situation--inadequate technical support for dealing with the arcane network operating system and a host of nagging hardware (in our case, initially cabling) problems. We were also motivated, however, by a larger pedagogic principle--indeed what has been our guiding principle in all our computer-based instruction--a commitment to training all first-year students and not just a select few, to dividing whatever computer resources we had (for our first three years, a single cramped classroom) among all first-year students (as many as 3200 a year). We were committed, in other words, not to develop classroom-specific computer practices--whatever we taught students in our classroom, we believed, students should be able to continue doing outside of class, at other sites on campus and off. Indeed, to ensure this universal access, we had students for years purchase and carry with them their own copy of an inexpensive but powerful word-processing program.

In the 1991-1992 school year, two changes occurred that made us rethink the role of networking in campus computing as well as a number of inherent limitations in personal word processing that we had not noticed before. The first change involved the proliferation of networks on our campus. By the end of the 1990-1991 school year, it was apparent that our insistence on students being able to load software wherever they worked from their own floppy-disk copy was soon becoming outdated: Our English computer lab was becoming one of the few places left on campus that was not loading software from a network server. At that time, I also learned of a key second change in which the university planned to connect all the campus labs together using broadband technology. No longer was our concern not to develop site-specific pedagogic procedures a problem because access to our server through the campus network meant that our own computer lab was, in a sense, expanding beyond its walls. With the campus network, students could access from anywhere on campus whatever specific services we might provide in our lab. Thus, there was no longer a good reason to avoid developing an electronic means for swapping files, facilitating first the most traditional academic exchange between student and teacher--handing out and collecting assignments--and after that what had already become standard practice in our writing classes--swapping papers among students.

We began thinking about and developing these file swapping utilities on the assumption that, as an extension of word processing, they would involve small, inconsequential, and readily achieved changes in what we were already doing. It all seemed simple: Swapping files would be the electronic equivalent of turning in, picking up, and swapping papers--or so we thought. Little did we realize at the time that in this single commitment to use networking capabilities for file swapping, we had taken a decisive step beyond the comfortable terrain of process pedagogy and personal word processing--we had, as it were, crossed a virtual Rubicon into a world where, given the right tool, students and teachers would really be able to work together collaboratively. This tool, we realized, was not yet developed, but we soon had a name for it--campus word processing. And over time, we developed a strong sense of how it must work, which would provide writers working together with as much power and flexibility as personal word processing had provided this past decade to writers working apart. What follows is a description, not of the campus word-processing system we have created at my own university--for we have yet to do so to our own satisfaction--but of the seven design principles we have developed in outlining that solution. Here, then, is a brief account of the basic design issues involved in developing an entirely new academic writing environment, one that conforms to Carolyn Handa's (1990) observation that the "world we learned to write in is not the world [our students] will be living and writing in" (p. 182)?

Design Principle 1: That campus word processing be accessible to all students campus-wide, that it not be restricted to students in certain classes or with access to a limited number of specialized sites

There is nothing wrong with using a restricted facility with a limited number of students to develop new programs and computer-based teaching techniques. Too often, however, such programs are developed as ends in themselves and marketed without any clear sense as to how more than a small minority of students on a single campus will ever gain access to these network facilities. Specifically, this first design principle continues to lead us in the direction of writing software that students may be trained on inside their writing class but which, like personal word processing, they regularly use to enhance their writing on their own, outside of any of their classes.

We thus remain committed to continuing what we have done the last seven fall terms--train all students inside their first-year composition classes for a new kind of computer-based writing that they will be using in various ways in different courses and settings throughout their college career. Because of this first design principle, we remain resistant to a major trend in the profession, the networked writing classroom as an actual place where students regularly meet their classes so that they can work online, sharing their current work and interests with each other. We remain committed, in other words, to the traditional classroom setting as the primary site of the essential oral and visual exchange between teacher and students. (Our university, like most, provides a maximum of only 37.5 direct, instructional contact hours between students and teacher--less than 2% of total hours in a typical semester.)

The network classroom remains for us a training site, a place where students meet on a short-term basis only to learn how to interact on the campus network out of class. We are committed, in other words, to asynchronous file sharing on a campus-wide basis, to transforming the entire campus into a virtual classroom--an electronic meeting place where students and teachers gather online at different times and different locations in order to share and respond to each others' work.

Design Principle 2: That campus word processing avoid student-generated filenames and other idiosyncrasies of personal computing

Our first solution to file sharing via network that would allow us to extend personal word processing to the campus as a whole entailed the development of what we thought would be a simple mailbox program. We would dedicate one or two computers in the foyer of our computer lab as places where students could turn in and pick up their work. Yet we soon discovered that even something as simple as turning in and picking up files via a mailbox program strains the limits of personal word processing in unexpected and seemingly trivial ways. Word processing, it turns out, is the computer tool par excellence of the individual writer, the person completing a writing assignment working entirely on her or his own. As a student in a class, one can respond to the directions of another, just as one can eventually either turn in the product of one's editing to the teacher or share it with classmates. All this sharing, however, ordinarily takes place via hard copy outside the word-processing environment. One shares a file by printing it, that is, after ending (if only temporarily) a word-processing session, and to insure the individuality of that copy, all one has to do is remember to include one's name.

The word-processing file itself is not so easily shared, in part because it is ordinarily named by individual users according to any idiosyncratic system they can devise. When we create or save a document, we have to give it a name--any name, although not presumably one that already exists. To turn in a document--what really entails saving it to someone else's disk--however, is an entirely different matter. Here, depending on exactly where one is saving a file on a file server and the user rights one has, an idiosyncratic filename is liable to overwrite someone else's file or an earlier version of one's own file. Conversely, one has to be careful that the file one is picking up also has a unique name and thus will not overwrite something on one's floppy disk. Although one can ensure a high degree of nonduplication by inventing strict protocols for naming files (protocols that assign specific functions to each of the eleven alphanumeric spaces in a regular DOS filename), one is still dependent on students rigorously remembering and following this convention. Simply stated, the file management conventions of traditional word processing--file naming especially--are inappropriate for group work.

Electronic mail provides a ready, even elegant answer to this second design principle. With e-mail, all the file management chores are performed by the program. When reading and writing messages, one never creates or directly manages files. Instead, one does something quite natural: One reads from and writes to someone about something. Over a period of time, one collects, not files named in some idiosyncratic fashion, but messages that can be arranged, sorted, or searched in common-sensical ways--by date, by sender, and, in more sophisticated programs, by key phrases. Yet e-mail, so good at managing files, nonetheless runs counter to the next three design principles.

Design Principle 3: That campus word processing be fully and transparently integrated with personal word processing

Here we are addressing two problems. First, we want anyone using the campus, networked system to have the full editing, formatting, and spell-checking capabilities of a full-featured word-processing program. Second, we want anyone temporarily without access to the campus network, and thus forced to use personal word processing, to be able to continue editing files already started on the campus system. Most current e-mail programs fail on both counts, lacking the full array of features of a word-processing program, and requiring a cumbersome import-export work-around, usually involving saving documents in ASCII format, in order to move text between two programs. With such ASCII work-arounds, valuable formatting information about margins and underlining inevitably gets lost--there is no way to avoid accidentally losing either formatting information or recent edits by trying to save both a current ASCII and a document version of each edited file.

Even the next step of being able to convert document formats in the background during the retrieve process, which is increasingly available with or in more sophisticated word-processing packages and e-mail files, is usually inadequate: One is still ordinarily faced with the task of editing the same file using two different sets of menus and keystrokes. Such file-filtering is acceptable only if the one of the two text editors is totally customizable so that it mimics completely the workings of the other. Such an elaborate work-around--involving both transparent file conversion and customizable menus--merely compensates for what should be an integral strength of a true campus word-processing system: a system that uses the same full-featured word-processing program, and hence file format, in both its networked and personal modes.

Design Principle 4: That campus word processing re-create electronically the normal reading-writing environment

Writing has always been supported by reading; whenever we write something using any technology, we always have the opportunity to read whatever it is to which we are responding. Traditionally, when we draft a response to someone's letter, we have that person's original letter in front of us. With word processing, we have assumed that what we responding to is alongside of us in printed form. In a campus environment where we are now reading online, we can no longer make this assumption. In order to get beyond the work-around of printing everything to which we are responding, we must design the screen so that we have access to both what we are reading (presumably on the top half) and to our response (presumably on the bottom).

Yet there is a critical problem here as well, one beside the limited size of the current 25-line computer screen. While it is not especially difficult to split the computer screen and allow one to read in the top while writing in the bottom, it is more difficult to make this process completely transparent, so that one can read and write together as easily on the computer screen as off. Indeed, such a transparent arrangement must allow writers working with text in the bottom window to scroll what they are reading in the top window forward or backward with the press of a single key--with any more complex combination of keys it becomes easier to print out and read from hard copy while writing online.

Design Principle 5: That campus word processing support the central role that the teacher has always played in initiating and mediating discourse within an academic course; phrased differently, that campus word processing be based on the fundamental difference between being in and out of a class

Both the strengths and weaknesses of e-mail derive from its being both flexible and unfocused: It is a system, specially designed, not for traditional educational use--that is, not to support the extended high-level literate exchanges between small groups of people focusing on a common topic that has long characterized the best academic communication--but like the telephone, for allowing anyone to be able to communicate with anyone else about anything. E-mail is designed to facilitate the exchange of usually brief messages between diverse individuals with some common, often momentary, concern, a not unimportant function, but one that in the past has had only the most minimal place in student-teacher interaction, and perhaps only slightly more in student-student collaboration. Alas, it is only preteens who attempt to use the telephone as a regular means of collaborative study; college students are more likely to use it to check with classmates and occasionally the instructor on missed assignments.

Campus word processing is based on a fundamental difference between the sustained and elaborated discourse that has long characterized academic writing and the short informal messages that characterize both telephone and e-mail exchanges. Although e-mail is inherently better than the telephone both for handling such messages and for facilitating group work because it allows one person (for example, the instructor) to reach many with a single message, like the telephone, it encourages people to pursue their own personal interests. Thus, in admittedly what some may see as an advantage, e-mail works against the efforts of the teacher or any individual student to focus the attention of a class or group as a whole, or even a subset of the group, on a single issue. What is at issue here is the difference between the free flow of information that is rightly the norm in informal social gatherings--in parties, for instance, or student conversation outside of class, both in their own ways invaluable sources of education--and other situations, including many (although not all) academic courses, where the lack of structure is less appropriate.

Everyone has sat through public meetings, for example, nominally called to act on a specific item of business yet that is conducted with neither formal agenda nor procedural rules--the meeting where each person in turn gets the floor to make a motion or more likely just offer an opinion that either refers to something already dismissed or broaches an entirely new subject (one that will likely be ignored by the next speaker). The issue is less whether or not a class is teacher-centered (or a meeting is chair-centered) but how good a job a teacher (or chair) does in facilitating discussion. The teacher, like the chair, is responsible, not for leading all discussions or making all motions, but for ensuring an essential level of orderliness--all class discussions, like all motions in meeting, as it were, flow through the teacher's (or chair's) implicit control.

It is finally not an either/or question, a matter of either giving or denying students access to e-mail--clearly students are entitled to a system that allows them to communicate freely with anyone they want out of class. The issue instead is in providing both students and faculty networking software that allows them to interact within a traditional course structure. An adequate campus word-processing system must be capable of offering teachers and students not just the ability to communicate with each other, as in e-mail programs, but to conduct and participate in an orderly, focused electronic meeting, as in the more sophisticated electronic conferencing programs that represent a significant step beyond e-mail. Such conferencing programs tend to compensate for the scattered, talk-about-anything-anytime quality of e-mail programs by allowing a user to create and invite people to join in a new, issue-specific topic. The structure of such a program promotes greater focus by mirroring the rule of order for conducting a meeting that requires the discussion to be in support of or opposed to the current motion.

Electronic conferencing as a key component of campus word processing is built on a recognition of the difference between the largely unmediated conversations that characterize being out of class--either via the telephone or at social gatherings--and the kind of directed discussion that characterizes more formal, issue-oriented gatherings, both traditional academic classes and formal meetings as well as the new kind of meeting that may be a better model for a campus word processing system: the three-part audience-expert-moderator discussion, pioneered by daytime talk shows such as Donahue and Oprah and now increasingly used in more serious evening news formats such as Nightline and the MacNeil-Lehrer News Hour. This new model of group discussion offers an alternative for teachers who both value greater student collaboration and are ready for something to replace the traditional role of teacher as always-present expert yet who are suspicious for a variety of reasons of the notion of relinquishing all control to the vague "will of the class." The goal here is to seek a middle ground between the absolutist and often off-putting statements of teachers and the quick and often unstudied impressions of students, to set up a third level text as "experts"--namely, the carefully worked out, measured responses of students. In this new model, the teacher becomes a moderator, facilitating a discussion between the class as audience and the invited guests--that is, the formal written responses of students here presented to the class as expert testimony. Such an approach to campus word processing necessitates the maintenance of a basic dichotomy between formal and informal student responses, between the type of measured thought and writing that has always gone into the best student writing and the kind of spontaneous, free-flowing response that has always defined the best informal class discussion. Here, campus word processing in supporting two modes of student writing described in the following design principle rejects the final weakness of e-mail (one shared with more issue-oriented conferencing programs), a flatness of structure that treats all written messages as equal.

Design Principle 6: That campus word processing be based on a basic hierarchy of texts, specifically between writing as the formal presentation of one's position (the traditional academic essay) and writing that helps one arrive at and revise that position (informal discussion and commentary)

Electronic conferencing programs are often no better than e-mail programs in maintaining this key distinction between our best written response to an issue and conversation, written or oral, that we might engage in both in developing and in revising that response. E-mail and conferencing systems tend to move our attention forward as we constantly respond to the latest message. "Less palpable and permanent than print," Michael Spitzer (1989) notes, computer conferencing, "like telephone conversation . . . feels conversational and temporary" (p. 193). Although no one can deny the value of such "talk"--written as well as oral--education has long given emphasis to a special order text, that which represents one's rehearsed and measured response. Indeed, without this notion of a dual hierarchy (what we are calling formal response and informal comment), the entire notion of the writing process essentially dissolves: Unless one is constantly rethinking and reediting one's text--that is, working backward as well as forward--there may be revision of one's ideas but there can be no real revision of text. Without a higher order text, one never gets past the stage of about-to-revise.

Design Principle 7: That campus word processing embody the basic difference between good collaboration (where people instinctively want to work with others because they see that to do so is in their own best interests) and bad collaboration (where people are forced by a higher administrative force to work together)

This final design principle emerges as an implicit response to one important question: Why do college students write so little, often even in literature classes and often despite the ready admission of their professors that writing is indeed such an instrumental part of learning that all students should be writing regularly in most of their class? A big part of the answer is that there is no one to read the literal volumes of material that dozens (much less, hundreds) of engaged students are capable of producing, almost on a weekly basis--and rightly or wrongly many teachers remain reluctant to require a lot of writing that no one is ever going to read. The typical two-part solution is (a) not to assign much writing and (b) to make up for it with discussions, either in class or in special discussion sections that meet in conjunction with large lectures.

The weaknesses of this system are obvious. First, we all know from our own experience that writing really does help us clarify our thoughts--that it is a powerful tool for critical thinking. All writers know the experience of coming to a deeper understanding of a topic as a result of having to organize and carefully articulate exactly what one wants to say. Second, it is difficult and expensive to schedule enough small discussion groups to ensure that everyone really has a chance to participate, and in many cases, there's no practical way to involve the professors themselves in most of the discussion sections that are scheduled in conjunction with their lectures. Finally, it is not clear just how small that discussion group has to be to ensure that all participants have the opportunity to make a real contribution in a 50-minute period, or just how skilled discussion leaders have to be to keep both themselves or a handful of overeager students from dominating the discussion.

The solution to this dilemma, as Cynthia Selfe (1990) notes, is to use the new technology to "invite more people into active engagements with, and conversations about, texts and encourage them to participate in different, and perhaps more egalitarian, ways than might be impossible using more traditional media" (p. 122)--and, one might add in meeting Design Principle 7, all without increasing current faculty workload. The solution, in other words, as we are attempting to do now locally, is to develop network computer software that in meeting the first six design principles allows students enrolled in the same class to meet either synchronously or asynchronously in small, virtual discussion groups. Here they can best do what, in the words of Ken Bruffee (1984), students do when working collaboratively on their writing: "Not write or edit or, least of all, read proof . . . [but] converse" (p. 645).

At the heart of this new campus writing environment is this simple four-part sequence. First, the teacher identifies an important area for consideration: an informal discussion question typically asked, often rhetorically, at the start of class; a more formal prompt for a quiz or more formal written response like an essay; or even something less teacher-directive--an open call for class feedback or comments, or an observation or suggestion that originates with a student. Second, students working at networked computer terminals, either inside or outside of class, draft responses to this prompt. Third, students working in small groups (a maximum of eight) share and comment upon each others' responses. Fourth, based on their reading and discussion, students revise their original response. To repeat: one, the teacher prompts; two, each student responds; three, the group discusses; four, each student revises. (Note: In order to support more traditional assignments such as tests, the teacher will have control of when, or even if, the responses are shared.)

Students collaborate in such an environment, not as is so often the case today, because the teacher tells them to read and critique each others' work (often just before it is ready to be submitted--and thus after almost everyone has already formed his or her opinions), but in an open and dynamic environment, indeed, in the very midst of thinking through and formulating their own response to the topic. This new network writing environment brings students together after their interest in the prompt has already been sharpened by their having completed their own initial contribution, even when that prompt focuses exclusively on prewriting. First respond, then share: "This is what I said; now let's see what others wrote!"--all at a time when, subject to the limits imposed by the instructor, students are still free both to sharpen and refocus their thinking and to revise their writing as they read their classmates' responses and exchange comments with them online. Campus word processing is thus an attempt to provide students with a controlled environment similar to a well-run class (or meeting), a place that provides participants greater opportunities to communicate deeply held and carefully articulated convictions with both their peers and with professionals trained and paid to encourage and respond to such efforts, that is, with their teachers. The goal is to create an environment with sufficient focus on one issue at a time that people online become what they too infrequently become offline: as passionate about other peoples' ideas as their own. The goal here is to follow Mike Rose's (1989) observation that students learn to write best when they have the chance to "write for people who care about language, people who are willing to sit with them and help them as they struggle to write about different things" (p. 54).

In the Information Technology section of a recent edition of the Chronicle of Higher Education, Beverly Watkins (1992) describes the successes and trials of Case Western Reserve University in installing and learning how to use a campus computer network. Although no one can question the immense advantage that such networks offer administrators and students in areas such as registration and scheduling and library circulation, their role in improving instruction is less clear. Indeed, what is especially impressive about Watkins' report is the seeming imbalance between the size and expense of the effort at Case Western Reserve (some $100 million over 10 years, with $40 million already spent) and the meager educational innovations they have yet realized. To be exact, she mentions four specific cases, all involving e-mail: (1) a classics professor from Case Western Reserve who communicates with a colleague from Purdue, (2) a night-owl physics professor answering e-mail in the early morning hours from his night-owl students, (3) three couples who first met on the network and later married, and (4) a group of students who were able to plan a surprise party successfully by communicating entirely by e-mail.

Surely no campus can justify such massive investment of funds for such meager educational benefits, even during the seeming rare periods of fiscal expansion. Yet this imbalance is finally the result not of spending too much on hardware but too little on software, specifically of assuming that traditional academic work, grounded as it is in the exchange of deep, reflective writing, can be enhanced by the computer equivalent of the telephone and its increasingly common counterpart, voice-mail systems. Case Western Reserve, simply stated, has been using the wrong tool, albeit perhaps the only one available to them. A true campus word-processing system, developed along principles outlined here, will represent a major advance beyond either e-mail or personal word processing, combining, as it were, the flexibility and convenience of the former with the explorative, expressive power of the latter. Students and teachers in such a world will still use each tool separately to meet a host of personal needs. At other times, however, when logged in to their class via the campus network, first pondering and responding to directions of their professors and then sharing and discussing their responses with their classmates, they will be working in a new writing environment, using a new writing tool, one designed to allow future students and teachers to reenact electronically the best pedagogic practice of the past: that passionate engagement with issues and language that transforms diverse individuals into a true community of learners.

Myron Tuman teaches in the English Department at the University of Alabama, Tuscaloosa.

Brodkey, L. (1987). Modernism and the scene(s) of writing. College English, 49, 396-418.

Bruffee, K. (1984). Collaborative learning and the "Conversation of mankind." College English, 46, 635-652.

Handa, C. (1990). Politics, ideology, and the isolated composer. In C. Handa (Ed.), Computers and community: Teaching composition in the twenty-first century. (pp. 160-184). Portsmouth, NH: Boynton/Cook.

Heim, M. (1987). Electric language: A philosophical study of word processing. New Haven: Yale University Press.

Rose, M. (1989). Lives on the boundary. New York: Penguin Books.

Selfe, C. (1990). Technology in the English classroom: Computers through the lens of feminist theory. In C. Handa (Ed.), Computers and community: Teaching composition in the twenty-first century (pp. 118-139). Portsmouth, NH: Boynton/Cook.

Spitzer, M. (1989, April 29). Computer conferencing: An emerging technology. In G. Hawisher and C. Selfe (Eds.), Critical perspectives on computers and composition instruction. New York: Teachers College Press.

Watkins, B. (1992). University hopes a campuswide network will help give it a competitive edge. Chronicle of Higher Education, A18-A20.