My introduction to word processing early in the 1 980s was hardly ideal. The computer was actually an unfriendly terminal connected to the University's mainframe. The keyboard included a key labeled RUBOUT, which (I thought) threatened to unleash the power of an assassin if I mistakenly pressed it. The word- processing program was a line editor, better suited to programming computer code than to writing and revising sentences and paragraphs. The manual? Let me just say that in retrospect it makes today's manuals look clear and helpful. More disappointing still was the teacher, who was just as opaque. A technical whiz, she managed to confuse me every step of the way and revealed little of what a powerful instrument the computer could be in the writing process.

Since that time, computer equipment and word-processing programs have improved and so has most training. Most graduate students and faculty in English departments have taken eagerly to word processing as a tool for their own writing. But what about training in the use of computers for teaching writing and literature? There is some evidence (Wresch, 1987a) that a significant percentage of elementary and secondary English teachers are receiving some training in their education courses. For the most part, though, college English instructors are receiving little if any training in what is an increasingly important part of composition teaching.

Clearly, as the use of computers in writing instruction continues to grow, there will be an increasing need for training teachers, both graduate students just beginning to teach and experienced colleagues. The literature on computers and composition instruction is filled with advice about ways to use computers in teaching writing, but the literature has included little about training teachers. This article offers some suggestions for what such training should include: basic pedagogical principles, analysis of types of software, advice on evaluating software, and some suggested readings. As much as possible, I offer advice general enough to be adapted to different lengths of training. I try to convey the same basic principles in a semester-long course about computer-aided composition instruction as I do in a one-day faculty development or teaching assistant training session. Obviously, longer training provides opportunities to flesh out these principles, to read, to discuss, and to practice. Shorter training mandates more prescription and less practice and exploration, despite their importance.

Good training in computer-aided composition instruction should, first of all, provide teachers with information-about possible uses of computers in writing instruction and about different types of computer programs. This information needs to be manageable in size and systematically organized. Second, teachers need practice using a computer and the programs they will use in their teaching. Although learning how to operate the computer should not be the focus of teacher training, without adequate knowledge of the computer's operating system and of file management, teachers encounter unnecessary frustrations when something goes wrong, as something always will. Teachers also need practice with programs in order to determine how to use them in their teaching. Program demonstrations and descriptions, without practice, are too ephemeral. Third, and most important, training needs to instill in teachers a necessary balance between an unbridled enthusiasm for using computers in writing instruction and an uncompromising skepticism about them. Bryan Pfaffenberger argues for such a balanced perspective, a kind of enthusiasm tempered with healthy skepticism: "[W]e must bring to computing the same questioning, critical attitude that we bring to our scholarship. That approach is the only way to overcome received wisdoms, tacit assumptions, and the other enemies of clear perception" (1987, p. 187). And as Douglas Sloan (1985) argues, there does not need to be a binary choice between being pro- or anticomputer. Instead, teachers need to choose which uses are appropriate, and which are not.

In order to make intelligent choices, teachers need sound guidance about using computers in writing instruction. Throughout the training I conduct, I try to emphasize the following:

1. Teachers should beware of both the positive and the negative hyperbole about computers in education. Some teachers uncritically accept exaggerated claims about the virtues of computers as writing and teaching tools and expect miraculous changes in their students' writing and in their lives as teachers. Other teachers fear or disdain the use of computers in teaching. On philosophical grounds, some see computers as anti-humanistic and thus antithetical to the teaching of writing and even to teaching in general. Others worry that the use of computers can mask bad pedagogy. On more practical grounds, some teachers fear that it is impossible to learn and keep up with ever-changing technology, and they object to spending valuable class time teaching students how to use a computer program. The truth lies somewhere between these extremes. Computers are not going to do all that the zealots proclaim. But neither do the machines necessarily ruin relationships between instructor and student or interfere with instruction. What makes for good writing instruction-frequent writing, discussion of writing in progress, revision, thoughtful responses from peers and instructors-does not need to change, indeed should not change, when computers are introduced to instruction.
   

2. Teachers should decide on educational goals and methods first, then consider how computers can be useful (not the other way around). Obviously, the capabilities of computer programs open up new teaching methods and help teachers achieve new goals. But I would argue that it is crucial for teachers to set their overall goals for a writing course before considering how computers can help achieve these goals. Otherwise, teachers run the risk of confusing the use of computers with teaching writing.
   

3. Teachers should consider computers primarily as tools for writers, not as omniscient teachers. The earliest uses of computers in education were built on an entirely different model than that which predominates today. Students were to learn about discrete writing skills from pre-programmed instruction. The computer controlled the pace of instruction and allowed little choice for students or instructors. Today, by contrast, computers are used most commonly as tools to help student writers do the things writers do as they compose-generate and organize ideas, create and edit text, and share drafts for comment. A surprising number of teachers new to computer-aided instruction in composition are unaware of this distinction.
   

4. Teachers should consider using computers as part of their instruction, not as the instruction. This advice has a variety of implications, some related to the points above. There is a natural tendency once one begins using computers in teaching writing to try to use computers as often as possible and to look to computers to solve every teaching problem. In fact, I would advocate that teachers start modestly, using computers in a deliberately limited way. Similarly, teachers who hold their class meetings in a computer lab should not feel obligated to have every class meeting there, or at least the lab should allow for instruction that does not involve computers. It should, for example, have tables for small groups to meet away from the potential distractions of printers and function keys.
   

5. Teachers should know that just because programs can do something does not mean it should be done. There are, as we all know, all sorts of bad instructional programs. Too often, teachers end up using programs built on principles they would otherwise never endorse; they use the programs simply because they are available. A good example of a problematic instructional program is the readability feature in many text analysis programs. Most students cannot interpret and use the information provided by these calculations (much less reconcile the differences among the scores produced by different programs).
   

6. Teachers should know that there is no Platonic ideal of a program for writing instruction. Of course, I am advocating that teachers should look critically at programs to make sure they are based on sound pedagogical principles and that they fit the course in which they will be used. But even when examined and chosen carefully, every writing instruction program has its flaws. Some word-processing programs lack some features one wants. Others have too many features and thus are harder to learn. Packages of pre-writing programs may include some options particular teachers do not like or that most students will not use. Unless teachers want to write their own software, they will need to compromise.
   

7. Teachers should be realistic about the time and the costs associated with using computers in writing instruction. The business model of computing implies that computers save time and money. That is why computers are so popular in business. Certain uses of computers, such as word processing, can save teachers time when they revise their handouts or their own scholarly writing. But using computers in writing instruction especially the first few times-consumes enormous amounts of time. Teachers find themselves learning to use programs, negotiating for class time in campus computing facilities, formatting student disks, writing manuals for programs, solving technical problems, and evaluating the success of their pedagogical experiments. If teachers are fortunate, someone else (an administrator or a pioneering colleague) has already secured adequate funding for computers, programs, and maintenance. Otherwise, teachers spend time they do not have to secure funds which are usually inadequate. No teacher contemplating using computers should be protected beforehand from these harsh facts.
   

In order to use software in writing instruction, teachers need to see a variety of possible ways to use computers and software in their classrooms. (I mean "see" literally here; actual programs should be demonstrated and teachers should have a chance to practice with them.) To make sense of this variety, teachers need a classification system for different types of uses. Literature about software offers numerous possible classification systems, organized around different principles. What follows is a system I find helpful in explaining the field to newcomers, some brief information about each type of program that I try to convey in training, and references for more information. What matters is not the particular system as much as having a system of some sort. The uses are arranged roughly in order of my sense of their importance; because I do not have time to introduce all of the types of programs, I start at the top and go as far down as I can. One last point- when I mention particular programs, I am suggesting examples of types. Certainly others could be substituted for demonstration and discussion.

Word Processing
This is of course the most common use of computers in writing instruction, and rightly so. Numerous practitioners, researchers, and theorists argue the advantages of providing students with the opportunity to write using a word processor and to have composition courses meet in computer classrooms. Word processing suits a writing curriculum that emphasizes revision. As Craig Etchison explains, the computer can become a writing tool that "can help foster those writing habits that have proven beneficial to experienced writers-habits such as revision, real revision, not just editing minor surface details." The computer can help "students understand the recursive nature of writing," by making it easy for them to respond to the developing needs of their papers (1985, p. 22). Although it may hardly seem necessary, it is appropriate to remind teachers that the mere use of computers will not foster the habits Etchison describes. Teachers should hear what Carolyn Boiarsky reminds researchers:

Computers do not revise; students do. Computers do not teach revising; teachers do. The word processor can function only as a tool to allow students to revise more easily, not as an instructor to help students acquire better revision techniques. (1988, p. 51)

Contrary to the expectation that writing classes meeting in a computer lab are sterile or mechanical in some way, such a setting seems to encourage a workshop atmosphere, one that facilitates talk about writing among students and the instructor. Teachers should be prepared for such a change in their classroom, and they should be encouraged to take full advantage of the computer as a teaching tool. They can, for example, quickly make multiple copies of a sentence or paragraph from a student's paper, have the student(s) experiment with possible revisions, and then compare the effectiveness of each version. In a classroom with a video network, it is easy to do such a demonstration for all students to see; without such a network, the same thing can be done on an individual computer for one student or a small group. Ronald Sudol (1985) and Barbara Waxman (1987) provide good advice about managing a writing course taught in a computer classroom and about teaching students to use a word-processing program to explore rhetorical options. Barbara Hutson and Diane Thompson explain the powerful effect computers and collaboration can have in a classroom:

If it is true that we learn about physical aspects of the world through handling objects, moving them around, it may also be true that we learn about language as a system in part by acting upon it, moving elements around and observing the effects upon meaning. When we write alone, the intuitions about choices of words and structure are private, ephemeral, and often below the level of consciousness. But when we compose and revise with others, linguistic awareness and strategies for problem solving in language are likely to be made public and relatively explicit. (1985, p. 57)

Pre-writing Programs
There is a wide variety of these programs, ranging from simple outliners to interactive tutorials (such as SEEN, discussed by its author in this collection). These programs guide a student through generating and organizing ideas for a particular kind of paper, and they can include thinking, invention, planning, and organizing systems. Sometimes, several prewriting programs are bundled together with text analysis and revision tools in an integrated package, such as HBJ WRITER or WRITER'S HELPER STAGE II. The latter makes available to students a variety of means, from formal to fanciful, to generate and organize ideas. Teachers should also know that it is possible to create prewriting tutorials without doing any programming at all; instructors can write questions in standard word-processing files and then copy them to student disks as instructional files or computer exercises. Timothy Weiss describes such a teaching technique in a recent article (1989).

Revision Tools and Text Analysis
Revision tools and text-analysis programs are designed to help writers revise prose written on a word processor. Charles Smith (1989) and William Wresch (1989) distinguish between those revision tools which help writers see their text in new ways (by collapsing an essay around key sentences, for example) and those which identify possible problems with spelling, word choice, grammar, and style. The most commonly used text-analysis program, and probably the most valuable, is a spelling checker. When they go beyond checking spelling, all text-analysis programs have serious limitations, limitations teachers should see first-hand during their training (there are, of course, limitations in spelling checkers as well). Al though there have been significant improvements in the newest versions, text-analysis programs frequently miss problems they are designed to find, identify errors where none exist, and offer inappropriate advice because the programs obviously cannot understand the text they are analyzing. Teachers should be cautioned to think carefully about whether their students will benefit from the analysis offered by these programs, especially whether their students are mature enough as writers to distinguish between those responses which are accurate and helpful and those which are not.

Communication
Communication programs offer exciting new possibilities for writing instruction. They use the computer as a type of mailbox and telephone to allow writers to do such things as brainstorm collaboratively, share drafts for comment, and engage in a written discussion on a course-related issue. Such programs are of two general types. The first, an electronic messaging system, is akin to regular mail-a writer posts a message for others to read and reply to at a later time. Composition instructors and their students have found effective ways to use such systems in their composition classes to distribute assignments, to pose questions, and more generally to extend class discussion beyond the boundaries of regular class meetings (Kinkead, 1987). Some writing instructors are using Bitnet, a wide-area electronic mail network, to allow students to exchange drafts for review with peers on other campuses in other states, thus trying to broaden the response students get to their writing. Another use of this electronic message capability can be found in the SEEN program's electronic bulletin board on which students post their responses to tutorial questions for other students to read and comment on.

The second type of communication program used in writing instruction is more like a telephone conference call, with students participating at the same time, although communication is through writing, not speech. Participants compose messages to each other on one half of a computer screen; completed messages appear as part of a running dialogue among participants on the other half of the screen. Such programs as ENFI (developed at Gallaudet University) and INTERCHANGE (developed at The University of Texas) are used as alternatives to oral discussion at various stages of the writing process, from brainstorming to revision. Teachers who use such programs find that they equalize student participation; that they encourage the precision in response that writing, as opposed to speech, demands; and that they create a new form of written discussion, which students enjoy.

Tutorials, Drill and Practice
This is a broad category, encompassing programs designed to teach skills ranging from semicolon usage to prosodic analysis. Although the computer may provide an efficient means to deliver tutorials and exercises on grammar and punctuation, having exercises on the computer rather than in a workbook does not change the fact that these types of exercises are of questionable pedagogical value. Teachers need to be reminded of this fact when they are introduced to these programs. If instructors want to use tutorial and drill programs, they should search for those which have clear pedagogical aims and accurate content; programs that give students immediate feedback on their work and that provide instruction as well as practice; programs that allow the instructor to change questions and answers; and programs that take advantage of the computer's graphic capabilities and its ability to tailor instruction to individual students through branching. Unfortunately, too few of these programs have such features.

Information Sources
Computers can open up worlds of information to students through electronic databases on all sorts of subjects, and composition teachers can bring this information into their classroom to help students learn about the process of doing research. Teachers can also have students create their own databases on subjects the students are researching, an excellent means for students to learn about selecting, sorting, and classifying information.

Authoring Programs
Authoring programs allow composition instructors who are not programmers to develop their own tutorials. Although these programs are still far from being as flexible and easy to use as English teachers would like, these programs make it possible for nonprogrammers to design much more sophisticated instructional programs than they could a few years ago. Teachers should be warned, though, that no matter how good the authoring programs, designing and developing instructional programs consume enormous amounts of time.

Once teachers determine what type of programs they want to use in their writing courses, they need to be able to evaluate particular programs. To help teachers choose programs and to influence program developers, the National Council of Teachers of English has published guidelines for evaluating programs. In addition, numerous books offer criteria for evaluating instructional computing programs in general and writing programs in particular. These criteria tend to be very specific and often do not indicate which items are most important. What follows is a brief and simple list of what I consider the most useful questions to ask about programs. For the most part, the questions apply to programs other than word processing packages. Most of the questions are self-explanatory, so I will present them without much elaboration. As was the case with classification systems, what is important is not so much which particular list of criteria teachers use to evaluate programs, but that they know that evaluation is necessary and that they have some means to do it. Obviously, instructors are often constrained in what programs they can use by budgets and by prior departmental selections. Nonetheless, they ought to ask tough, critical questions about the programs available to them, and they should feel comfortable choosing not to use certain programs.

1. Does the program teach something worth teaching? This is, of course, what I have been advocating as the fundamental question about using computers in writing instruction. As Wresch puts it,
   

   Is the whole thing worth the bother? If a program is easy to use, is accepted by students, and covers a concept you intend to teach, then it will be used. If the program keeps breaking down, is rejected by students, or teaches a concept no one has taught since 1957, it won't be used. (1987a, p. 112)
   

2. Is the program based on sound principles of how people write and how people learn? Research, theory, and practice have told us much about how writers write and about how students learn to become better writers. It is appropriate to expect computer programs designed to teach writing to be informed by these ideas. However, does the program actually encourage students to think, to create, to write? If we believe that students learn to write by writing, by struggling to find and make meaning in a text, good instructional programs should have students engaged in actual writing, not merely selecting answers for multiple choice questions
   

3. Is the program flexible--for the student and for the teacher? Does the program give students choices about how to use it? If the program is a tutorial, for example, does it allow students to start in different places? To stop before the end? To resume where they left off? Does the program allow the instructor to change the questions and responses? Helen Schwartz's SEEN is a model in this regard; it comes with a number of pre-writing tutorials, but it also allows instructors to create their own, without having to be programmers.
   

4. Is the program easy to use? The hard work should be done by the programmer, not the user. There is, of course, a trade-off between ease-of-use and sophistication or flexibility; but even sophisticated, flexible programs should be relatively easy to use.
   

5. Does the program offer collaborative possibilities? Many of the best uses of computers in writing instruction provide opportunities for students to cooperate and to learn from each other, rather than to be isolated from one another.
   

Readings
Obviously the time available for training teachers in computer-aided composition instruction will determine how much reading about the field they can do. There are now, fortunately, a number of good books and articles to recommend. To warn teachers of some dangers and to calm their fears, all teachers using computers in a writing course for the first time should read Barbara Waxman's delightful essay "Catching the Runaway Train" (1987), which describes her experience teaching her first computer-aided composition class. For an overview of the types of programs available and of ways they can be used in teaching composition and literature, Wresch's The Computer in Composition Instruction: A Writer's Tool (1984) remains valuable. Wresch's Practical Guide to Computer Uses in the English/Language Arts Classroom (1987b), though designed for elementary and secondary teachers, offers much of value to college faculty as well. Hawisher's articles (1986, 1987, 1988) as well as Bridwell-Bowles' survey in this issue, reporting on research about the effects of word processing on writing, help make sense of a complicated subject that demands attention in teacher training. Within the past two years, a number of books have been published that offer a more critical perspective on the field, all of these readings can generate lively discussion in training sessions. Among them are Davis (1988), Dobrin (1989), and Hawisher and Selfe (1989).

I purposely do not provide a prescription or detailed syllabus for training, because the amount of time available for computer training and the user needs vary so much from one English department to another. Regardless of the time available for training, I would argue that writing teachers should learn about the kinds of programs available, the possibilities for instruction in computer-aided composition instruction, and the principles for using this instruction suggested here. Those of us involved in computer-aided instruction must work to develop effective training programs as part of professional development activities for faculty and graduate students. Like other pedagogical training, this training should be carefully planned and well-taught. This training should also be ongoing, in order to provide teachers with a chance to share their problems and successes with colleagues. The goal of this training should be to create knowledgeable instructors who have a critical perspective on what they do. Stephen Marcus points to the importance of informed instructors as we move forward in using computers in writing instruction:

If anything can be predicted about computers and literacy with much assurance, it is that some people will oversell the technology as a positive force and that others will decry it as an abomination. Perhaps the most we can hope for (and it has always served us well) is a continued reliance on talented teachers acquiring an informed exuberance. As always, they will be a major force in making the most of whatever the technology and their students offer. (1985, p. 116)

Bradley T. Hughes is Director of the Writing Lab at The University of Wisconsin at Madison.

Boiarsky, C. (1988). Determining the effects of word processing on composing: Another perspective. Computer-Assisted Composition Journal, 2(2), 51-57.

Davis, S. (1988). The ragged interface: Computers and the teaching of writing. In F. V. Bogeland & K. K. Gottschalk (Eds.), Teaching prose: A guide for writing instructors (pp. 337-392). New York: Norton.

Dobrin, D. N. (1989). Writing and technique. Urbana, IL: National Council of Teachers of English.

Etchison, C. (1985). Who's making the decisions--people or machines? Computers and Composition, 2(4), 17-26.

Hawisher, G. (1986). Studies in word processing. Computers and Composition, 4(1), 6-31.

Hawisher, G. (1987). The effects of word processing on the revision strategies of college freshmen. Research in the Teaching of English, 21, 145-159.

Hawisher, G. (1988). Research update: Writing and word processing. Computers and Composition, 5(2), 7-29.

Hawisher, G., & Selfe, C. (Eds.). (1989). Critical perspectives on computers and composition instruction. New York: Teachers College Press.

Hutson, B., & Thompson, D. (1985). Moving language around on the word processor: Cognitive operations upon language. Quarterly Newsletter of the Laboratory of Comparative Human Cognition, 7, 57-64.

Kinkead, J. (1987). Computer conversations: E-mail and writing instruction. College Composition and Communication, 38, 337-341.

Marcus, S. (1985). The future of literacy. In D. Chandler & S. Marcus (Eds.), Computers and literacy (pp. 108-116). Philadelphia: Open University Press.

Pfaffenberger, B. (1987). A review of the wordworthy computer. Computer-Assisted Composition Journal, 1, 185-189.

Sloan, D. (1985). Introduction: On raising critical questions about the computer in education. In D. Sloan (Ed.), The computer in education: A critical perspective (pp. 1-4). New York: Teachers College Press.

Smith, C. R. (1989). Text analysis: The state of the art. Computer-Assisted Composition Journal, 3(2), 68-77.

Sudol, R. A. (1985). Applied word processing: Notes on authority, responsibility, and revision in a workshop model. College Composition and Communication, 36, 331-335.

Waxman, B. (1987). Catching the runaway train: Computers and the advanced composition class. Computer-Assisted Composition Journal, 1, 208-213.

Weiss, T. (1989). A process of composing with computers. Computers and Composition, 6(2),45-59.

Wresch, W. (Ed.). (1984). The computer in composition instruction: A writer's tool. Urbana, IL: National Council of Teachers of English.

Wresch, W. (1987a). Preparing the next generation of English teachers to use computers. The Computing Teacher, 15(3), 41-43.

Wresch, W. (1987b). A practical guide to computer uses in the English/ language arts classroom. Englewood Cliffs, NJ: Prentice-Hall.

Wresch, W. (1989, March). A history of computer analysis of student writing. Conference on College Composition and Communications, Seattle, WA.