In The Four Loves, C. S. Lewis observed, apropos of sharing a friend with a third party, that "Far from seeing less of A when B is around, I am given a particularly A perspective of B which I should have missed otherwise." Lewis was a tutor himself, and a good one according to an acquaintance of mine who was his student. Lewis is describing an interpersonal dynamic that is at the heart of the tutoring situation. In my experience as a writing center director, the computer often plays the role of this third party, becoming an instrument that concretizes intentions, abilities, problems, and conflicts of people engaged in teaching and learning the writing process. Keeping this relation in mind, I am going to discuss what for me are the four most important areas in which computers affect the process of tutoring other writers: tutor-to-client dialogue, macrostructural revision, surface editing, and printing. Rather than citing statistics about how a given function helps an average client, I will focus on the human dimension through four informal case histories that illustrate ways in which computers contributed to the tutoring dynamic for an English major, a Chinese student minoring in writing, a high school teacher, and a doctoral candidate. These cases support many of the conventional generalizations often made about the effects of computers on composition, but the cases also make clear that the computer has some unexpected catalytic effects on human interactions. These effects are especially obvious in writing centers, where instruction focuses not only on the subject matter but also on human relations.

Writing centers deliver "personalized instruction," that is, instruction that takes the personality of the learner into account. The basic assumption of personalized learning is that the learner is a normal person with learning preferences that tutors ought to respect, insofar as is consistent with good instruction. Writing center tutors are taught to see problems in the context of the relationship between the student and the classroom teacher and to assume that both teacher and student have good reasons for their beliefs and behaviors. "Personalized instruction" may mean taking the person's writing seriously, or giving the person permission to deviate from a teacher's agenda, or helping a person to adjust an unproductive attitude, or coping with academic emergencies of various kinds.

The basis of personalized instruction is the tutor, the third person who sees both the student's and the classroom teacher's perspectives, and who becomes a medium of communication between student and teacher. In the following four cases, the computer also shares the role of the third person, mediating between client and tutor as the tutor mediates between client and teacher. The first case illustrates how the computer can supply a social basis for the one-to-one relationships upon which personalized learning is founded, and the second case illustrates how the computer can increase the flexibility of one-to-one instruction. The third case illustrates the point that, for teachers, who are often among the clientele of writing centers, the computer can make clear the triviality of grammar drills and practice exercises and help shift the focus of instruction from proofreading to more productive areas of the writing process. The final case illustrates how the computer can help to solve a perennial problem of writing centers: writers who appear in writing centers needing tutorial assistance late in the writing process.

Anyone who has entered a word-processing classroom recognizes the change the machines bring about in the social structure of the class. Given a common task and a common language, such a class coheres much more quickly and tightly than do non-computer classes. The same non-computer phenomenon occurs in the writing center's one-to-one environment. A good example from my early days of tutoring on computers was Marcie, the first student I taught to use word-processing software.

Before 1982, when we acquired microcomputers in the writing center, we tutored some students on an IBM mainframe word-processing system called ATMS, for Automated Text Management System (or some such thing). It was a Godzilla, with indecipherable documentation, no menus, and tutorial screens intended for rocket scientists. I had learned to run ATMS from Tom, a linguist who worked in the computer center and whom I continued to call when I had a question. I was surprised when Marcie, a C-average senior English major with no particular skill as a writer, asked me to teach her to write on the computer. Marcie was a part-time receptionist for the writing center, so I showed her how to sign on--no small task because it involved entering secure codes for the writing center's account--and then I let her type for half an hour until she was ready to save. She wanted to see her work printed out. I explained that output had to be sent to the computer center across the street where it was queued up with output from every office and computer science class on campus. Marcie decided against printing, and I thought that would be the end of it. The next day, however, she was back, and, in trying to sign on, she made an error. Her cries for help were audible in the hall outside the writing center, and thereafter I grew accustomed to hearing my name sung out plaintively when she encountered a problem with the system. She became as reliant on me as I was on Tom, my computer guru and source of ATMS advice. As I demonstrated word-processing functions, I would insert REM comments (nonprinting notes) in the text or add little snippets that made transitions or illustrated a point. Sometimes I would hit the CAPS-LOCK key and write a paragraph arguing with one of Marcie's. She would argue back, and some of her arguments were included in final drafts. Our exchange was fun. I only wish I could report dramatic improvements in her writing.

It became clear that she was neither successful nor happy majoring in English, and her student teaching confirmed that. She eventually became a hotel manager in St. Louis, where her charm and energy made her quite successful. I don't know whether she still writes at all. However, writing on the computer gave her an experience she had in no other context in college. Marcie was neither technologically inclined nor verbally proficient, but the machine conferred to writing an importance that Marcie enjoyed. Even more important, the situation gave her permission to ask questions. As an English major taking classes with people who were fluent writers, Marcie had felt it simply wasn't appropriate to ask for help on writing. However, as an English major, she was permitted a certain amount of cyberphobia, and in the context of asking questions about the machine, she requested and I supplied writing advice. We talked about writing, we took her projects seriously, and she completed them. Several conferences with me and her Colonial American Literature professor yielded a paper which he gave a B-. She had gotten Bs before and might have gotten the B- anyway, but she had never before been interested enough in what she was doing to confer with two professors over a paper. Whether another student would have reacted that way, or whether Marcie might have been a better English major if she had learned word processing as a first-year student, is not so important as the fact that learning to run the computer was the occasion of her learning to ask for help.

When I introduce the computer to new tutors, I explain that the dialogic relationship between tutor and client is the basis of effective one-to-one instruction. The computer supplies a social basis for that relationship because it represents a common interest and a new language with which to discuss that interest. The computer also provides motivation in that achieving small successes is easy, and each success leads to further interest in achievement. In addition, the computer facilitates the tutor's commentary, demonstration, and dialogue on student papers.

In this case, I was the referring teacher, not the tutor, for a student I worked with in a writing class during the spring term of 1986. Eileen, my student, was from Hong Kong. She spoke English fluently and was not at all the stereotypical Oriental student who dutifully completes every detail of assigned tasks. Rather, more like many American students, she had a habit of procrastination that she covered with an ever-present smile and a tactic of changing the subject or asking a disarming question whenever she was cornered. I ran the Writing Seminar, the capstone course for people majoring or minoring in writing, as a studio, relying on students to arrive with drafts of manuscripts for discussion and mutual criticism. This class met twice a week for 75 minutes in a classroom outfitted with 18 Zenith microcomputers, each machine connected to a Citizen printer.

Eileen took to the computer naturally. Before the course began, she had already been to the University Center for Learning Assistance (our old writing center, newly expanded and reorganized that year) and had learned to sign on, type, save, and print. My introductory demonstrations were wasted on her, so I showed her how to reform paragraphs, how to delete whole words and lines with a couple of keystrokes, and how to run the spell checker. Teaching her these and other word-processing procedures always required only a matter of moments. My problem was persuading her to write papers.

Eileen was a person who felt deadline pressures very late, and her dilatory habits were not well-suited to my plan for the course, which presumed motivated students with writing agendas extending over the entire semester. At the first workshop session during the second week, the students were to look at each other's drafts. Eileen showed up with only a couple of paragraphs. Because she had little to offer, she commented on the draft of another student in the class, and then sat at the computer looking through the menus. I sat down next to her with a disk I had prepared for a class activity later in the week. It contained the first draft of an essay on Arthurian legends. I told her the essay needed revision and asked her to suggest some changes. She deleted some lines and added a few.

Because she had already established a relationship with a tutor in the Center for Learning Assistance, I suggested that Eileen continue to work on the Arthurian legends essay in the Center. I hoped that in the course of revising the essay she might work more on her own writing. In fact I was wrong. She kept attending class with partially completed assignments and finally turned in most of the work during the last couple of weeks of the term. The portfolio was written well enough that I felt I had to give her a B, though I wanted to give her a C for her exasperating work habits. Her tutor, a male graduate student in English named Chuck, informed me that she had scored as "apprehensive" on the Daly-Miller scale. His records of three one-hour sessions with her indicated that, because she was apprehensive about writing, some of his time had been spent showing her how to use WORDSTAR's BLOCK functions to manipulate large chunks of text. The discourse-level rearrangements and deletions of whole paragraphs she completed under Chuck's tutelage indicated that she could conceptualize a whole essay even though she seemed reluctant to produce one.

Both Chuck and Eileen seemed to see the computer as a toy. Given a piece of text, they played cut-and-paste with it as if it were a video game. They also "played" with Burns's invention program and a student statistical package called MICROSTAT. Because my positions as teacher of Eileen's writing course and administrator of the Center for Learning Assistance forced me to take both computers and writing a good deal more seriously than they did, I did not officially sanction the video-game orientation. But my years of experience in learning centers have taught me to restrain my tendency to interfere in tutoring: Tutors and clients have a good sense of what is important for them to do together. In this case, Chuck's judgment that Eileen needed to relax seemed as sensible to me as my own desire to see her produce on schedule.

It is rare good luck if the agenda of a course happens to coincide with the developmental imperatives of a person. If I begin thinking that, as a teacher, I make all the difference, it is useful to remind myself that people became formally operational long before Piaget was there to give them problems in the conservation of mass. In combination, the writing center and the computer are especially powerful forces to realize developmental goals. Many centers dissipate that power by reliance on computerized versions of dull handouts and programmed learning modules. But for clients who are naturally motivated to explore text structure, word-processing menus present a list of options available to writers, thereby prompting experimentation. The CUT and PASTE functions concretize basic writing processes of addition, deletion, substitution, and rearrangement. Invention software facilitates outlining, brainstorming, freewriting, and other forms of creative thought. Statistical packages provide a means of analyzing information unavailable to centers without computers. Chuck and Eileen took advantage of all of these offerings, and they did so at their own pace.

During the summer of 1986, I taught a man named Andy, who was enrolled in my four-week workshop designed to train teachers to use computers in the classroom. I didn't get along with Andy, and I didn't teach him much, but he did learn something from the course, and I give the credit to the machine. An English teacher in a traditional department of a large suburban high school, he confided to me before the workshop began that he was a process-oriented teacher. I was skeptical because I knew his department chair to be a teacher preoccupied with clean copy and famous for translating each surface error into a point value appropriate to its gravity. Andy knew that writing centers were places where remedial students ran drill-and-practice programs to learn the principal parts of verbs, so he was sure that I would be able to suggest some programs that would solve his students' problems with standard written English. I confessed that I wasn't aware of any such software. However, I told him, if such a program ever came out, I was sure it would be in all the newspapers because it would be the answer to every teacher's prayer.

In the first five minutes of the workshop, I saw Andy's face assume a frozen smile when I gave the participants my cardinal principle of tutoring with word-processing software: My rule, which I call the Code of Silence, is "Make them drag word-processing information out of you." He clearly didn't agree when I advised the participants that if they wanted to focus their writers on the task at hand, they should not lecture on operating the machine. He pointedly did not take notes when I urged them, jointly and severally, to spend the first hour teaching writers to sign on, type, save, and print a personal letter, and then to discuss word processing per se only in response to direct questions during one-to-one tutoring.

Andy thought that a word-processing system should work like a food processor: You throw some language into a bin, and it comes out a neat, uniform product suitable for immediate consumption. In those days, our word-processing systems, unfortunately, were ancient 64K Columbias that ran WORDSTAR 3.3 rather slowly and that locked up when you tried to cancel a print. The Columbias were on tables in rows arranged so that I couldn't see the faces of the 20 students if I stood in the front of the room, and I could scarcely get between the rows to tutor them. Andy mentioned each of these inconveniences during his first half hour in the workshop and made them the subject of entries in his journal. Because he had some experience with Apples and APPLEWRITER, he made invidious comparisons between the Apples and the Columbias, between APPLEWRITER and WORDSTAR. The operation of the DELETE command seemed particularly to exercise Andy. It was to him among life's bitterest ironies that one had to place the cursor on the letter in order to delete when, in his view, God had manifestly intended us to backspace over errors.

Andy was threatened throughout the early going in the course by the exercises I gave they all involved writing rough copy that others could view, taking risks, and thinking in the chair by, for example, taking turns writing paragraphs that argued a debatable point. He warmed up a bit when I showed him how to run the spell checker; clearly this was familiar terrain. He was annoyed that the machine couldn't find repeated words or misspelled words that formed other correctly spelled words. He thoroughly enjoyed a pairs exercise I put the students through in which each member of the pair was to teach the other some subtlety of a WORDSTAR menu. Clearly he felt successful discoursing on the Quick Menu to the woman who sat next to him. He steepled his fingers and spoke in a voice audible all over the room.

Late in the workshop, I instructed the participants in the use of RIGHTWRITER, a simple style program that identifies sentence length, repetition of words, and so on. When Andy examined the copy of his document that contained the comments inserted by the program, he triumphantly showed me where the program had mistakenly identified some linking uses of "to be" as instances of the passive voice. What good was the program if it made errors? My explanation, that it couldn't actually process syntax because artificial intelligence researchers were still working on a parser, bemused him. He seemed reluctantly pleased by the idea that at some distant time the machine would produce text-editing comments. I demonstrated how, in the meantime, he might treat the output as heuristic in nature. He could use the mistakes the machine made as exercises to test his students' understanding of principles of style and mechanics.

I gave both precept and example of the teaching practices that my experience suggested would create the most focus on the writing process: giving out word-processing information only as it was needed, emphasizing the creation of a sense of community among the writers in the group, subordinating surface editing to discourse-level revision, and expecting no miracles from the computer. But Andy saw the machine in the context of the political system in his fiscally and educationally conservative institution. If he was to justify spending money or time on the computer, it had to involve a new set of terms and procedures to teach and test. The syllabus he created for his workshop project listed lecture time for menus and spell checking, but none for in-class conferences or drafting. GRAMMATIK n and RIGHTWRITER also figured prominently, but he hadn't included any of the pair exercises. I was annoyed that he evidently hadn't heard much of what I had said, or seen much of what I had tried to model. But as I thought about his using the spell checker and style programs, I realized that he had undergone one profound change.

The change was in his attitude toward surface errors. He had come to the class looking for drill-and-practice programs that would, through some Pavlovian dynamic, train students to spell and punctuate. His agenda didn't match mine at all, but to some degree, the computer itself arbitrated our disputes. If a program could do some but not all of the proofreading in a paper, then the proofreading skill was both less necessary and more subtle than he had realized. Andy left the class willing to delegate much of proofreading to a spell checker or style program, in fact dissatisfied that present software wouldn't do the whole job.

I deal daily with teachers who, like Andy, equate writing with proofreading, who want me to reassure them that tutors will not proofread and thus "do the work" for the clients. Computers help me to help these teachers place the problem in a more balanced perspective. If a writer can use software to check spelling and style, proofreading becomes about as important a writing skill as typesetting. In the long run, true grammatical analysis of sentences may become possible. In the meantime, style programs such as GRAMMATIK II assist in identifying sentence problems. Tutors and clients may discuss and evaluate the output of style programs, treating these programs as heuristic devices rather than merely accepting changes. In the presence of editing software, grammar drill-and-practice programs are reduced in importance, becoming media for instruction in language as content, not as skill. Writing instruction is focused on more productive areas.

In a sense, computers have revived a consciousness of the product in composition. The technical advances of software and hardware in desktop publishing have given every writer the ability to produce text that is difficult to distinguish from typeset copy. Inevitably this technology encourages some mindless tinkering with typefaces and pointless printing of rough drafts. However, it has also given writing centers the means to face up to a responsibility that writing instruction has too long ignored. The mechanical aspects of text production have traditionally been considered pre-college instruction, even though "clean copy" may mean markedly different things from one context to another in college. With computers, the contributions of page layout, graphics, and typeface can be brought into the tutorial situation.

Nick is a doctoral student with whom I have worked off and on for the past six years. Nick is a fluent writer and speaker with a tendency to take the easy way too often with his graduate work and to finagle political solutions to the problems he creates for himself. He took longer than most students to complete his course work and comprehensive examinations, and he had to retake parts of his examinations several times before passing. He then wrote and received approval from his committee for his dissertation proposal. Much more rapidly than I had anticipated, he composed, at the computer, a writer-based first draft of his dissertation that received negative reviews from his committee members, along with a host of suggestions for changes. His second draft received more positive reviews, but some of the objections to the first draft were repeated, and the scheduled defense was delayed. The first draft had been printed on a dot-matrix printer, and the second draft on a good-quality daisy wheel printer. At this point, Nick began asking for explicit instructions for changes, feigning helplessness in order to pressure committee members, or so it seemed to me.

An interesting political problem began surfacing in the committee. Some of the members decided that Nick was trying to slip substandard work past the committee. One member in particular wrote comments that attacked basic premises of the dissertation approved at the proposal stage. I can't give too many details. Suffice it- to say that the situation deteriorated rapidly as committee members began taking sides. Finally the dissertation director took a more active role in arbitrating the dispute, a number of changes were called for and made, and the dissertation was, in the end, successfully defended. Revised, the dissertation was good enough to stand on its own, but before the defense, Nick's advisor made a shrewd suggestion that helped diffuse the tension and make the defense the collegial rite of passage it is supposed to be. Nick's dot-matrix copy had not helped his case with the objecting committee members, two of whom were accustomed to printing their own work on laser printers. The daisy wheel second draft looked about as good as ordinary typed copy; however, this situation called for extraordinary measures. Nick's advisor suggested that Nick get the copy for the defense printed on a laser printer, which he did . The defense was not without some moments of controversy, but the professional appearance of the final copy helped both Nick and the committee members to see past the earlier forms and to take some satisfaction in the changes. The extra effort of shifting to the professional mode of printing was an important symbolic gesture in the context of Nick's tendency to take shortcuts.

Without the computer, the whole process would have taken many more discouraging months to complete. Similar interactions occur frequently in writing centers. Like Nick, clients who see the center as an intensive care unit often appear in writing centers desperate for last minute assistance, throwing themselves on the mercy of the tutor. Clients in this condition need help, not lectures on improvidence. Because of computers, those writers who need to revise can receive tutorial assistance late in the writing process, sometimes minutes before the paper is due. When the words and ideas are right, the printing is no problem. Thus, word processing and the printer lead to less anxiety about the appearance of the final product because it equalizes writers' abilities to produce aesthetically appealing text. A clean draft becomes no great achievement, enabling tutors and clients to focus on issues of substance.

Robotic factories and science-fiction novels have contributed to the impression that computers are a de-humanizing influence on the environment. My experience is exactly the opposite. I have experienced computers as a humanizing force in that they impart to writing some of the naturalness, intimacy, and flexibility of oral language situations while retaining the elegance of the printed word. As Lewis says, "We possess each friend not less but more as the number of those with whom we share him increase" (p.92). To "I" and "you," the computer adds "he," "she," or "it," a third force that is neither person, but which lets the first and second person see more of each other. As the cases illustrate, the effects of computers are as various as the personalities and intentions of the people who use them. The one-to-one environment highlights the rich diversity of the learning experience sometimes obscured in teacher-to class environments. It seems clear to me that, as more writing is taught in computer laboratories, the flexible one-to-one models of instruction developed and practiced in writing centers will gain wider acceptance. In these new contexts, the computer will operate like a good analyst, providing a blank screen onto which we can project who we are and who we want to be.

Maurice Scharton teaches at Illinois State University.

Lewis, C. S. (1960). The four loves. New York: Harcourt Brace, Inc.