7(3), August 1990, pages 109-127

Instructional Software for the Introductory Composition Classroom:
One Teacher's Review and Recommendations

Debbie D'Aoust

Why instructional software for composition? Over the past few years, a number of software packages have been developed that use computer technology to assist writing instructors with many of the difficulties they face: presenting models for the processes of writing; getting students to begin writing and to think about what they are doing; providing help when it is most needed (while the student is actually writing); providing feedback on a timely basis so that problems in the written product may be linked to problems in the writing process; being available for frequent consultation on a one-to-one basis; and maintaining patience and a pleasant attitude when repeating the same information over and over again, either to a single student or to a number of different students. This paper reviews and makes recommendations concerning eight such software packages designed for undergraduate settings. The programs to be reviewed include the four writing division "Distinguished Software" award winners from the 1987 EDUCOM/NCRIPTAL Higher Education Software Competition: PROSE, SEEN, and SUPER SCOOP I and II; three other submissions to the competition: EDITOR, PROTEUS, and WRITING IS THINKING; and one well-established commercial program: HBJ WRITER.

In order to establish a consistent base for evaluating the software and because it is a situation facing many writing instructors, I posited a scenario in which I was the teacher for one section of an introductory composition course. Conversations with both teachers and students confirmed that students in this course would be likely to have a wide range of abilities and experiences with writing. At the University of Michigan, students and teachers have access to the Apple Macintosh and Zenith computers in the campus computing centers; I chose the undergraduate library computing center (a very lively place at all hours of the day) as my research site. Two of the programs I review (SUPER SCOOP I and II) are not available to run on our hardware, making them less suitable for class use. In software development, the IBM PC is becoming the standard (Bump, 1987); but at the computing center, there were always long lines to use the Macintoshes, while many Zeniths sat unused. This situation is a significant indicator of student preference that should be considered in further development and purchasing plans. It is, perhaps, not so much a preference for the Macintosh as for what the Macintosh computer can do. Graphics often play an important role in student writing.

There are many complex, interrelated issues that must be addressed when evaluating any composition software, but a central question is "How do you think writing should be taught?" (Miller & Burnett, 1986, p. 163). Evaluation criteria concerning accuracy of content are perhaps more easily judged in software for math or the sciences than for the field of composition, where there seem to be as many competing theories about teaching as there are people involved in the field. Nevertheless, the question must be answered. Although it is no doubt preferable to state one's beliefs before beginning the review process, my own answers emerged largely during the review process as I encountered a number of different strategies and reacted to them. The two assumptions that became central in my evaluations are that writing models should be structured enough to be helpful but flexible enough to permit creativity (a very delicate, and perhaps impossible, balance); and that both teacher and peer evaluation are beneficial.

Although in theory I agree with Holdstein (1987) and Bridwell and Ross (1984) that a variety of "courseware" should be available to students, my scenario suggested that I would, at least initially, have limited funds available to make such purchases. This factor introduced the pressure to orient my evaluations towards singling out the one or two programs that would aid the largest number of writing processes, appropriate for students with the widest range of writing abilities and experience, and that would be useful in connection with the greatest variety of assignments. There is, perhaps, no such thing as an all-purpose program for writing instruction, but my evaluations are in large measure an attempt to identify those that most nearly approach this ideal.

EDITOR

EDITOR by John Thiesmeyer stands with WRITER'S WORKBENCH as one of the most powerful text editing programs available. It accepts ASCII word processing files and uses pattern matching with four usage dictionaries, which together, the documentation states, can locate "over 10,000 common problems with word choice." Fix locates definite errors such as doubled words (although note, in the sentence "It's amazing that that speeding car didn't crash," the doubled that would not be an error). Tighten identifies locutions with excess verbiage and is based on the principle that shorter is better. Polish identifies clichés and other stale usages, and is based on the principle that fresher is better. Consider locates cases calling for user judgment. EDITOR also contains a vocabulary analysis program. This program will display a frequency list and mark those words, except common function words, that occur more than four times on a page, with the suggestion that more variety might be needed.

Unfortunately, every time I tried to view the output of a program run on the screen (a routine operation), the screen display would become garbled, and I would be dumped back into DOS. In addition, neither the on-screen directions nor the printed documentation were complete for the version of EDITOR I was testing. Even if these conditions had been met, however, I do not believe I would recommend EDITOR for use in my classroom. I say this even though similar editing programs received a positive response from students because these programs reinforced the marking of texts by teachers. The programs also received positive responses from teachers who found that their grading time was spent more fruitfully (Kiefer & Smith, 1984). It is not, in my opinion, a beneficial program for beginning writers.

A program such as EDITOR seems ideal for confident writers, but I must agree with Holdstein (1987) that such a detailed analysis of "errors," with pages and pages of commentary, may be intimidating to basic writers (p. 125). Another problem is noted by Thiesmeyer (1987), EDITOR'S author, who characterizes his program's potential as "no small thing" (p. 113), but states that in his experience students have an even lower tolerance for errors in pedagogic software than for its human counterpart. Thiesmeyer further argues that once students discover that the software is fallible, they often begin ignoring its editorial promptings. This finding coincides with my own experience with spell-checking programs and other editing programs. Yet another drawback is that these programs deal largely with matters of style and are not equipped to deal with ungrammatical texts (Hull, Ball, Fox, Levin, & McCutchen, 1987). After surveying the current state of artificial intelligence and natural language understanding research, Thiesmeyer (1987) concludes that we will not be able to process artificially such things as sentence fragments, comma splices, and malformed sentences, let alone matters of coherence and argument, within the immediate (or probably even the distant) future.

SUPER SCOOP I and II

The absence of natural language understanding also detracts from the programs SUPER SCOOP I and n by Peter Owens. These programs model the journalistic experience of following up on a story lead through a series of interviews. And they are engaging exercises in gathering and evaluating information, offering excellent incentive to practice note-taking skills. Yet, I found the interview process inadequate and frustrating in each program. In SUPER SCOOP I, the student is passive during this process, except in choosing whom to interview, because the program provides both the student's questions and the other party's answers.

In SUPER SCOOP II, students have the opportunity, at certain points during certain interviews to formulate their own question. This input is checked for a very limited number of key words, independent of the context of the question; and if any are present, a set answer is provided. If none are present, the student is given a chance to ask a second question; and if no match is found, the fictive character responds with a message that he or she does not have time to "talk" any longer, and the interview is terminated. On only one occasion was I able to provide a keyword that matched, and, in that case, the answer provided talked about my word ("money"), but did not answer my question. Of course, this might say as much about my journalistic skills as it does about the limitations of current dialogue-inquiry CAI. Both of these programs were awarded "Distinguished Software" citations by EDUCOM/NCRIPTAL, and I think they have their place in a large software library. However, for my scenario, they are too narrowly focused (the actual writing of the story is not done within the program) and would have too short a shelf-life (being appropriate for only two, or three at the most, uses by any one class) to warrant placement on my "priority list."

PROSE

Another program that received an EDUCOM/NCRIPTAL "Distinguished Software" citation is PROSE (Prompted Revision of Student Essays) by Stuart Davis, Nancy Kaplan and Joseph Martin (Cornell University). This software facilitates softcopy exchange of papers and comments among students and teachers. PROSE allows a teacher to open a text at any point to insert comments that the student may read (or ignore) and erase, or to insert specific prompts that will detain the student until a change (any change) is made. Although such a system does have certain attractions, problems of access time may arise when the entire writing, marking, and reviewing cycle is done at the computer.

I would not make PROSE a priority for my fictional classroom because it does not address the pre-writing or writing activities of the student and because the capacity for inserting commentary in a text is a feature of HBJ WRITER, which also performs a number of other functions. (See later section on HBJ WRITER.)

PROTEUS

A similarly limited software package is PROTEUS by Robert Leonard, who calls it an "idea processor." It consists of five pre-writing activities, the results of which may be printed or converted to an ASCII word-processing file for development into a paper. The five pre writing activities are Freewriting, Looping, Listing, Five Ws, and Cubing. Freewriting, an activity in which the computer will beep if the student stops writing for more than ten seconds, aims to overcome writer's block by encouraging students to get ideas out without worrying about matters of structure. Looping is a routine in which the student may go back to the freewriting text and input more writing at any point (but not edit the existing text). Listing is like freewriting, but is done in list form and can also be used in conjunction with looping. The last two activities encourage more structured idea development. Five Ws asks students to list ideas about their topic in response to who, what, when, where, and why questions. Cubing presents students with six points of view from which to view their topic (Analyze, Compare, Contrast, Argue For, Argue Against, and Define) and asks the student to freewrite about each. This input can then be further developed using any of the first four exercises.

Because of its open-ended structure, PROTEUS avoids the problems of other idea processors or outliners that restrict creativity to the words already on the page and make symbol manipulation a substitute for thinking (Dobrin, 1987). Within PROTEUS, students are not restricted to what they have written because the program discourages editing and encourages more writing, which must be printed out or transferred to a word-processing program for re-working. As a teacher, I think students' time would be well spent if they used this program to explore their topic fully before they began to write; but as a student, I think the program suffers from its insistence on the separation of pre-writing and writing activities. I found myself frustrated at not being able to get on to the business of putting my ideas together in some less fragmented and more coherent manner. In addition, I found that the pressure of the freewriting exercise (with its insistent and embarrassing beeping) was ultimately more inhibiting than liberating, and in the ensuing panic, I began just pounding out meaningless strings of characters to satisfy the machine. Some of PROTEUS's activities would no doubt be valuable to some students at some times, but I can not recommend it as a priority purchase.

WRITING IS THINKING

A program that does integrate pre-writing, writing, and revision, and that attempts to address the "meta-skills" of purpose, audience, coherence, voice, and so on (Bums, 1987), is WRITING IS THINKING by Mary Kappel. Figure 1 shows an overview of the program's structure. This interactive tutorial, which guides the student through the development of a short opinion paper, clearly falls into the category of "heuristic programs that reinforce the student throughout a well-established 'process' for writing by means of a series of programmed promptings" (Barrett & Paradis, 1988, p. 156). Such programs are open to the criticism that they may be too "behavioristic" (Whiting, 1986), because their underlying "Skinnerian" premise is that students will eventually internalize the prompts, so that "the user is being programmed as well" (Barrett & Paradis, p. 156).



[Figure 1]

Figure 1: Overview of WRITING IS THINKING program structure.


Whether or not we put much weight on such criticism will depend, it seems, on whether or not we approve of the behavior that is being programmed. It seems unlikely that attempts to encourage students in habits of examining different points of view and questioning their own assumptions would come under fire. The question to be raised with WRITING IS THINKING is whether its rigid structure actually brings new insights to students'' thinking, or whether it just forces useless patterns of thought (Dobrin, 1987). The program's assumption that paragraphs are formed by inserting supporting facts after a sentence that expresses an opinion runs the risk of producing the 'lifeless and formulaic" content seen in the standard five-paragraph essay (Bridwell & Duin, 1985, p. 116) and raises the concern that if instructional programs do not allow enough flexibility for beginning writers to evolve as diversely as successful writers have, "we may inhibit some important creative activities" (p. 119). As a student, I found the approach to paragraph development restrictive and overly determinative; and to a large degree, I would agree with Dobrin (1987) that the greatest benefit of such programs may be that in watching the way they interfere with writers' purposes we can learn something about what writers are trying to do (p. 106).

These objections have to do with the delicate balance between being structured enough to be helpful and being flexible enough to permit creativity, and are largely confined to the two central Develop the Structure sections of the software. I found the Analyse the Problem section of WRITING IS THINKING to be an effective tutorial on developing a question for a paper to answer. I do not, however, agree with its premise that the opinion or argumentative paper is the basic form of writing under which all others must be subsumed. I also believe the promptings in this program overlook the subtle difference between choosing a question that has only one answer and having a single goal for a paper.

The Evaluate the Design section is, I believe, WRITING IS THINKING's strong point. This section assists teachers in their efforts to help students see the value of deep revision as opposed to mere rewording (Curtis, 1988; Hult, 1988) because it separates the processes of revising and editing. In both areas, the emphasis is on asking students questions to consider, thus stressing student thought over software sophistication. In the area of editing, this method also avoids the pitfalls of traditional programmed editors that attempt to identify the faulty or questionable constructions within the text. Some of these potential problems were mentioned earlier in connection with EDITOR Another potential problem is that students may come to rely on the machine to do their thinking for them and neglect errors that the software cannot catch. Yet another problem is the concern of Hull et al. (1987) that "calling attention to forms that are correct as possibly incorrect may confuse the writer unnecessarily" (p.106). The section on determining audience is also quite good; and throughout the entire program, drill and practice on key concepts is available in sub-routines that students may choose to bypass, allowing them an element of control in the program's operation.

Despite the several aspects of WRITING IS THINKING that are, in my opinion, superior to similar elements in other software, I am unable to recommend its use because of basic conflicts I have with the rigidity of its sections on developing structure. I do think it might be useful for certain students for a limited number of assignments, but the scenario I have laid out dictates that I select programs that are more widely useful. WRITING IS THINKING also suffers from the lack of an integrated word-processing program to which students could move to work further on their papers outside the constraints of the tutorial. The files created by the tutorial user can be transferred to an ASCII word-processing file to be used in conjunction with another program, but such a step is time consuming and distracting.

To the larger question of whether or not some version of a tutorial such as WRITING IS THINKING may be useful in composition instruction, the answer must be that such software has great (if sometimes unrealized) potential. The general consensus is that students enjoy the private nature of the interaction (Selfe, 1984), and at least one empirical study based on several different measures of testing revealed that CAI tutorials can be as effective as their human counterparts and more effective than classroom instruction (Gillis, 1987). Such findings should be encouraging because, as Burns (1984) notes, composition teachers do not always have time to "play Socrates to every student's Phaedrus" (p. 26). Available conference time between students and teachers may also be spent more fruitfully if routine dialogues are conducted on the computer tutorial.

HBJ WRITER

Another software package that integrates pre writing, writing, and revision, although in very different ways, is HBJ WRITER. Figure 2 shows an overview of the program's structure. Unlike WRITING IS THINKING, HBJ WRITER incorporates a full word-processing program that can be used on its own or along with any of the pre-writing or revising and reviewing aids. This makes it very appropriate for a class such as the one I have posited because writers of varying abilities can pick and choose features according to their own needs. In addition to standard word-processing features for moving through, arranging, and searching text, HBJ WRITER also has a split-screen feature. This feature is very useful because students can work on two versions of the same file at one time, or on two different files, such as their paper and either an outline or teacher instructions.


[Figure 2]

Figure 2: Overview of HBJ WRITER program structure

All of HBJ WRITER's pre-writing aids create files that may then be worked on by writers using a word-processing program. The first pre writing aid, Freewriting, is like the one in PROTEUS, but with some important changes that make it more useful. Students type approximately 150 words (the program counts seven characters as one word), during which time, if they stop typing for more than a few seconds, the message "Keep Typing!" flashes on the screen (infinitely more tactful than beeping). It is possible to exit gracefully (i.e., without turning off the machine) from this activity should a student wish to back out. After the freewriting is completed, students are prompted to take a moment to read over what they have written and to summarize the main idea it embodies or suggests. This cycle may be repeated as often as desired. Another pre-writing aid is Invisible Writing. Invisible Writing is also designed to aid in overcoming writer's block. In this routine, students must type 100 words before they can see what they have written. They may have another piece of text or outline on the screen as they write.

The other two pre-writing aids that HBJ WRITER offers prompt students to identify (in Nutshelling) their title, purpose, audience; and (in Planning) their thesis or main point, supporting statements, opposite viewpoints, and key words or phrases, along with an order in which they could be used . Both of these routines operate by presenting a small amount of information (for instance, Nutshelling states that a title should attract the reader's attention and tell what the paper is about) and then by asking students to complete statements such as "What I really mean to say in my paper is . . . " HBJ WRITER thus lacks the detailed guidance and review of concepts that WRITING IS THINKING offers again, there is a trade-off between structure and flexibility.

The Organizational Review section of HBJ WRITER contains a second version of Nutshelling. This second version is essentially the same as the first, except that it begins by asking students how long it has been since they last worked on the file and by suggesting a break of at least half a day. The prompts are exactly the same in both versions; however, in the second version of Nutshelling, the prompts are followed by the suggestion that the answers be used to review the paper and ask questions about the clarity of purpose, the suitability for audience, the obviousness of main points, and the logical relationship between these elements. The student may search his/her paper for either transitional words and phrases (using a database of 249 entries), or pronouns (using a database of 48 entries), or both; and these will be highlighted (or muted on the amber monitor I used) in the text. The program gives no indication about what the student should do with these highlighted words, so it would be up to the student's teacher to give instruction on the importance of these features of language for writing cohesion.

A third prewriting routine allows students to create an outline by extracting either the first sentence of each paragraph (which would only work if the paper had been constructed along the lines of the WRITING IS THINKING format) or by selecting sentences from throughout the paper. This outline-after-the-fact seems to be a good idea, as many students are too impatient or uncertain about what they want to write to construct an outline prior to writing.

In the Stylistic Review section, papers may be searched for to be verbs, prepositions, abstract words, -tion and -sion words, or "gender-specific nouns" (words like postman, not words like father). Matches for these words appear highlighted (or muted) in the text at the top of the screen, while the bottom of the screen shows the text marked with dashes for all unmatched words and letters for matched words (a series of as for abstract words, ps for prepositions, and so on). Then Stylistic Review presents statistics concerning the number and percentage of matched words and where they occur, and presents a judgment or comment. Samples of such "reports" to the student are the following

The responsibility for interpreting and acting on these comments rests with the student, and will largely depend on the specific context and any instruction given by the teacher. For instance, the comment that a paper does not contain many abstract words might be considered positive feedback for many types of writing, but such a comment would probably be considered negative if it were applied to a paper on critical theory.

The Mechanical Review section of HBJ WRITER contains a standard spell-checking program as well as databases to search for commonly misused words and punctuation errors. Possible errors are highlighted in the text, and the student can call up an explanation for why each entry has been marked. Some of these explanations are more useful than others. For instance, in the sentence "If I could turn the whole thing around and look in at the other end, why then the figures would . . ." the marked "why" was explained by the note that if this is a question it should end in a question mark. This explanation seems adequate in the example, although problems with non-errors being marked as possible errors have already been discussed. In another case, the explanation was more unsatisfactory. The word by with the author's name (after the title) was marked, and the explanation was to consider whether buy meaning "to purchase," bye meaning "goodbye," or by meaning "near or in" was intended when none of these matched the intended sense of "composed by." If there is a positive argument to counter these negative remarks about editing programs, it is perhaps that made by Ross (1985), who states that because they posit the writer "as an expert consultant who will approve, modify, or reject any suggestions" (p.109), students will have a stake in overriding the program and therefore will pay attention and learn something. This reaction may, however, depend on the student's capabilities.

Finally, HBJ WRITER contains a routine for commenting on papers. This routine allows a text to be opened at any point and any amount of text (which is automatically indented and highlighted) to be inserted. Comments may be signed (the default is "Anonymous"), and as many as nine different commented-on versions of a paper may be saved . This aspect of HBJ WRITER means that the comment feature may be used not only by the teacher, but also by the students when they exchange papers among themselves for potentially valuable peer analysis (see also Halpern & Liggett, 1984).

HBJ WRITER was originally developed as WANDAH (Writing Aid AND Author's Helper, by Ruth Von Blum, Michael Cohen, Morton Friedman, and Earl Rand) for use in first-year English classes in U.C.L.A.'s Writing Program. HBJ WRITER then underwent extensive testing at U.C.L.A., at the Universities of Minnesota and Washington, and at Panahou High School in Honolulu before being made available through its commercial publisher (Von Blum & Cohen, 1984). This testing has paid off, I believe, in HBJ WRITER's smooth operation, integration of functions, and concern for detail. Although not flawless, it seems to strike a reasonable balance between structure and flexibility, and it offers a number of possibilities for long-term uses in the varied setting I have assumed in my scenario. For these reasons, my own favorable review may be added to others HBJ WRITER has already received (Bump, 1987; Arms, 1987; Bridwell & Ross, 1984).

SEEN

Some teachers, like me, find the idea of a CAI tutorial to be an excellent one, but see WRITING IS THINKING as seriously flawed. Some of us may also hold the more general belief that '~ am the only one who knows how to teach my course, and no one else's programs are going to be any use to men (James, 1986, p. 268). The answer for us may lie in teacher authorable tutorials such as SEEN (Seeing Eye Elephant Network) by Helen Schwartz, which was awarded a "Distinguished Software" citation by EDUCOM/NCRIPTAL. It is no longer necessary, as it was in the earlier part of this decade (Davis, 1983), for teachers to develop the ability to program in computer languages because of this and other available authoring programs. Although more open-authoring systems will provide the capability for complex branching patterns, the one in SEEN has the advantage of being programmed to transfer the "idea files" created in the tutorial to an electronic bulletin board. On this bulletin board, the files may be commented on by teachers or other students, allowing for different types of evaluative processes.

The authoring program in SEEN (despite operational problems on the version I was testing and poorly written, incomplete documentation) demonstrates that from a set of four different types of frames (information only, limited response questions, open response questions, and multiple choice questions), each simple in itself, a sophisticated tutorial program may be constructed. As a starting point, I would recommend Burns's (1984) method of using taped sessions of tutorials to identify precisely what usually goes on in such exchanges. Alternately, teachers can begin by customizing SEEN's tutorials on critical reading or literary character analysis.

Conclusion

Why instructional software for composition? One possible reason is that given by Bums (1987)--that computers will influence composition instruction for better or worse, and that teachers must be knowledgeable to choose the right course. This answer actually implies two possible motivations, one negative and focused on defense, the other positive and focused on offense. Personally, I prefer the positive approach, for it is clear that computer technology is capable of, and is already, helping many composition teachers to deal with some of the problems outlined in the introduction to this paper. Published discussions of some of the larger issues concerning incorporating CAI into composition classrooms include Cohen's (1987) satirical, yet serious, account of the problems involved in evaluating the benefits of software to justify the expenditures involved; and Holdstein's (1987) comments on problems involved in software development ranging from programmer cooperation to tenure considerations, and the claims by some instructors and administrators that computers will replace human teachers. In regard to this last issue, I must agree with Thiesmeyer (1987) that given the subtle structures and contents of the language we teach, and given the current prospects in artificial intelligence and natural language understanding research, students and teachers of writing must remain "indefinitely dependent . . . on their own marvelous and mysterious linguistic abilities" (p.114).

This does not mean, however, that computers may not supplement the instructional process in various useful ways. The programs described in this paper all have the potential to benefit writing teachers and their students. For the introductory composition classroom new to CAI, however, I would specifically recommend HBJ WRITER and SEEN. Taken together, these programs cover the full range of writing processes and offer great flexibility for use in different types of individual or group assignments. Because the programs do not rigidly promote a specific philosophy of teaching, they should also be useful for paralleling or supplementing a wide variety of traditional texts and teaching methods.

Debbie D'Aoust is a teaching assistant for the Department of English at The University of Michigan.

Software Information

EDITOR
Serenity Software
901 Serenity Road, Penn Yan, NY 14527
IBM PC (Macintosh & Unix versions in preparation)
Site License $49.95

HBJ WRITER
Harcourt Brace Jovanovich
1250 6th Avenue, San Diego, CA 92104
Phone 800-237-2665
IBM PC
Teacher package $66.50; student package $26.50

PROSE
Kinko's Academic Courseware Exchange
4141 State Street, Santa Barbara, CA 93110
Phone 800-235-6919; 800-292 6640 (in CA)
Macintosh 512, 512e, Plus
Instructor module $10.00; student module $15.00

PROTEUS
Research Design Associates, Inc. / Mind Builders
P.O. Box 848, Stony Brook, NY 11790
Apple IIe, IIc; IBM PC; Tandy 1000; Commodore 64; TRS 80
Single copy $59.95; site license $99.95

SEEN
Helen Schwartz, Department of English
Carnegie Mellon University
Pittsburgh, PA 15213
Phone 412-268-5636
IBM PC
6 program/group disks ~ teacher disk $198.00

SUPER SCOOP I & II
COMPress, A Division of Wadsworth, Inc.
P.O. Box 102, Wentworth, NH 03282
Phone 800-221-0419
Apple II family
$65.00 each

WRITING IS THINKING
Kapstrom Inc.
5952 Royal Lane, Suite 124, Dallas, TX 75230
Phone 214-369-1718
IBM PC; Apple IIe
$165.00

References

Arms, V. M. (1987). Engineers becoming writers. In L. Gerrard (Ed.), Writing at century's end: Essays on computer-assisted composition (pp. 64-78). New York: Random House.

Barrett, E., & Paradis, J. (1988). Teaching writing in an on-line classroom. Harvard Educational Review, 58(2), 154-171.

Bridwell, L., & Duin, A. (1985). Looking in depth at writers: Computers as writing medium and research tool. In J. L. Collins & E. A. Sommers (Eds.), Writing on-line: Using computers in the teaching of writing (pp. 115-121). Upper Montclair, NJ: Boynton/Cook.

Bridwell, L. S., & Ross, D. (1984). Integrating computers into a writing curriculum; or, buying, begging, and building. In W. Wresch (Ed.), The computer in composition instruction. A writer's tool (pp. 107-119). Urbana, IL: National Council of Teachers of English.

Bump, J. (1987). CAI in writing at the university: Some recommendations. Computers and Education, 11(2), 121-133.

Burns, H. (1984). Recollections of first generation computer-assisted prewriting. In W. Wresch (Ed.), The computer in composition instruction. A writer's tool (pp. 15-33). Urbana, IL: National Council of Teachers of English.

Burns, H. (1987). Computers and composition. In G. Tate (Ed.), Teaching composition. Twelve bibliographical essays (rev. ed.) (pp. 378-400). Fort Worth: Texas Christian University Press.

Chignell, M. H., & Lacy, R. M. (1988, October). Project Jefferson: Integrating research and instruction. Academic Computing, 3(2), 12-45.

Cohen, M. E. (1987). In search of the writon. In L. Gerrard (Ed.), Writing at century's end: Essays on computer-assisted composition (pp. 116-121). New York: Random House.

Curtis, M. S. (1988). Windows on composing: Teaching revision on word processors. College Composition and Communication, 39, 337-343.

Davis, J. C. (1983). English teachers can create their own instructional software. Focus: Teaching English Language Arts, 9(3), 126-129.

Dobrin, D. N. (1987). Some ideas about idea processors. In L. Gerrard (Ed.), Writing at century's end: Essays on computer-assisted composition (pp. 95-107). New York: Random House.

Gillis, P. D. (1987). Using computer technology to teach and evaluate prewriting. Computers and the Humanities, 21(1), 3-19.

Halpern, J. W., & Liggett, S. (1984). Computers & composing: How the new technologies are changing writing. Carbondale: Southern Illinois University Press.

Holdstein, D. H. (1987). The politics of CAI and word processing. In L. Gerrard (Ed.), Writing at century's end: Essays on computer-assisted composition (pp. 122-130). New York: Random House.

Hull, G., Ball, C., Fox, J. L., Levin, L., & McCutchen, D. (1987).Computer detection of errors in natural language texts: Some research on pattern matching. Computers and the Humanities, 21(2),103-118.

Hult, C. A. (1988). The computer and the inexperienced writer. Computers and Composition, 5, 29-38.

James, E. B. (1986). Computer-based teaching for undergraduates: Old problems and new possibilities. Computers and Education, 10, 267-272.

Kiefer, K., & Smith, C. R. (1984). Improving student's revising and editing: The WRITER'S WORKBENCH system. In W. Wresch (Ed.), The computer in composition instruction. A writer's tool (pp. 65-82). Urbana, IL: National Council of Teachers of English.

Kollmeier, H. H., & Staudt, K. H. (1987). Composition students online: Database searching in the undergraduate research paper course. Computers and the Humanities, 21(3), 147-155.

Miller, L., & Burnett, J. D. (1986). Theoretical considerations in selecting language arts software. Computers and Education, 10(1), 159-165.

Ross, D. (1985). Realities of computer analysis of compositions. In J. L. Collins & E. A. Sommers (Eds.), Writing on-line: Using computers in the teaching of writing (pp. 105-113). Upper Montclair, NJ: Boynton/Cook.

Selfe, C. L. (1984). WORDSWORTH II: Process-based CAI for college composition teachers. In W. Wresch (Ed.), The computer in composition instruction. A writer's tool (pp. 174-180). Urbana, IL: National Council of Teachers of English.

Thiesmeyer, J. E. (1987). Expert systems, artificial intelligence, and the teaching of writing. In L. Gerrard (Ed.), Writing at century's end: Essays on computer-assisted composition (pp. 108-115). New York: Random House.

Von Blum, R., & Cohen, M. E. (1984). WANDAH: Writing-Aid AND Author's Helper. In W. Wresch (Ed.), The computer in composition instruction. A writer's tool (pp. 154-173). Urbana, IL: National Council of Teachers of English.

Whiting, J. (1986). Student opinion of tutorial CAL. Computers and Education, 10(2), 281-292.