1. Although the collaborative writing assignment may be announced and distributed on the very first day of class, it should not be begun until a substantial portion of the term has elapsed. In the interim, pedagogy should be sufficiently collaborative (e.g., collaborative class discussion, small-group work, collaborative invention, collaborative revision) that the students get to know each other, resolve some of the small interpersonal tensions that inevitably arise, and anticipate each other's collaborative assets and shortcomings.
2. The collaborative writing assignment should be one that is best accomplished by a group rather than an individual; otherwise, the task is artificial, leading to students' frustration and irritation. Lunsford and Ede describe

. . . three types of tasks which invite . . . collaboration: "labor-intensive" tasks that need to be divided into smaller subtasks in order to be accomplished effectively and efficiently; "specialization" tasks that call for multiple areas of expertise; and "synthesis" tasks that demand that divergent perspectives be brought together into a solution acceptable to the whole group or an outside group (Singular Texts 123).

In my sophomore composition class, the collaborative groups were assigned to write a new chapter for a style textbook used in the course. Because the students had criticized its shortcomings, I directed each group to write a new chapter for the book. One group wrote on wired style; another, on the ways style varies in different discourse communities. These were not topics that solitary sophomores could have treated authoritatively, for they demanded intensive labor and high-level synthesis.

3. Provide for student-initiated collaboration. While planning the course, consider whether any of the assignments designated for individual authorship might lend themselves to collaborative authorship. Alert the class to these possibilities and introduce them to methods and rationales for collaboration. If some of the students opt to write collaboratively, work with the groups to assure that they are accomplishing something other than dividing up an individual paper among several writers.
4. Discuss methods and problems of collaborative writing before the project begins. If online collaboration is a possibility, explore available methods (see Forbes; Selfe). Alert students to the ways in which stereotyped role expectations (based in factors such as gender and race) can affect the distribution of power within the group; encourage students to use the collaborative experience as an opportunity for greater attention to each other's ideas, and for delegating responsibility according to the actual characteristics of the individuals in the group, rather than their stereotyped role expectations (see Fox; Morgan; Villanueva). Two additional observations are especially important to share: (a) Collaborative writing can take many forms. Lunsford and Ede describe dialogic and hierarchical collaboration: in dialogic collaboration, the group works together on all aspects of the project, whereas in hierarchical collaboration, the group divides the task into component parts and assigns certain components to each group member. Lunsford and Ede point out that these are not mutually exclusive categories; many collaborative writing projects involve both dialogic and hierarchical collaboration (Singular Texts 133-134). Dialogic collaboration offers the benefit of discovery: students learn more by working together. In my class, the group working on wired style collaborated dialogically to define and expand the unfamiliar topic. Hierarchical collaboration offers the benefit of efficiency: that same group worked dialogically to develop a preliminary bibliography and then worked hierarchically to read the sources. Each member reported his or her findings to the group, and then they again worked dialogically to construct an argument and hierarchically to draft sections of that argument. (b) Certain problems regularly arise in collaborative writing. One is that some students are better prepared to accomplish their tasks than are others—hence a variation in quality. Writing groups need to be prepared for each person's contributions to be revised and even deleted. The group must be ready to exert critical judgment, and the members must be braced for the sobering prospect of having their work changed or eliminated. The group must dedicate itself to the best possible written product, and its members must be ready to help each other through potentially ego-deflating moments. (c) Another common problem in collaborative writing is the student who does not carry his or her load. The group should deliberate this possibility at the beginning of the project and report to the teacher its decision for how such a phenomenon should be handled if it arises. Generally, this involves the group's deciding whether one grade will be assigned to the paper regardless of the balance of effort; whether a shirking member will receive a lesser grade than the others; or whether a shirking member will be ejected from the group and either given a zero or required to write his or her own paper.
5. Anticipate and prepare for student resistance to collaboration. In part because Western literary theory has, for the past century and more, so firmly endorsed a model of solitary authorship, some students have difficulty accepting collaborative writing assignments. They may be uncertain as to whether their classmates will accept them as co-authors, or they may have a much higher opinion of their own writing "ability" than their classmates have. Whatever the cause of their resistance, most of these students will nevertheless benefit from the collaborative assignment, if they are given to understand how prevalent collaboration is in workplace writing; how much their "individual" writing will benefit from having worked in a group and having seen firsthand how others articulate and solve writing problems; and how much more they can accomplish than if they were working alone—what Linda Hughes and Michael Lund call "a union that is greater than the . . . parts that composed it" (49). (Priscilla Rogers and Marjorie Horton detail the benefits of collaborative writing, especially in the dialogic mode.) But some students are implacably opposed to collaboration, and the teacher must decide whether to require them to participate or whether to offer the option of individual writing. The decision rests on the pedagogical motivation for assigning collaboration. If it is to improve the students' skill in writing collaboratively, they must all participate. If, however, the class is designed to enhance individual writing skills (as is the case in most required introductory composition classes) and a student persists in objecting to collaboration, the teacher may want to allow solitary composition, reasoning that the student's writing skills will not benefit from an activity that he or she so firmly resists.
6. Let the class decide how the groups will be constituted, and discuss the pros and cons of each possibility. I told my class that choosing their own groups would allow them maximum comfort but would leave some students feeling unloved, and I also told them that the comfort of self-chosen groups could sometimes result in poor decision-making, with too much consideration for established relations and not enough for the collaborative project. Our alternatives were my designating the groups, trying to fix each group with a range of writing skills—e.g., a good researcher, a good editor, etc., in each group. The class decided on random selection, and they also chose the size of groups. I recommended against too-small groups (in which one person's absence might be devastating) or too-large groups (in which leadership issues could too easily arise and in which one person could too easily disappear). They decided on six-person writing groups.
7. Give the groups autonomy in deciding their methods and timetables (my class even decided how often, when, and why class would meet during the month of the project), but also require that they commit their timetable to writing. Give students maximum guidance to help them make sound decisions. The sophomores were sobered, for example, when they realized that once the paper was constructed, they would have to allow not a few hours but a week for editing, since they were all required to read and sign off on the final product. And indeed, for one group, editing was the most lively and protracted passage of the project. One student was determined that the paper be well edited and was certain he knew what good editing was. The other members were less enthusiastic but more knowledgeable about editing. The result was fights, dictionaries, handbooks, delegations to my office for grammar arbitration, a great deal of learning, and a beautifully edited paper.
8. Prepare for dissent within the groups, and prepare to manage it in two dimensions: the teacher and the students. Neither should attempt to suppress dissent or enforce consensus (see Clark & Ede; Flower; Janangelo; Spellmeyer; Trimbur; Villanueva; and Wolf). Successful collaboration, say Lunsford and Ede, allows not only for "group cohesion" but also for "creative conflict" and the protection of "minority views" (Singular Texts 123). Flower recommends that the teacher welcome rather than dread dissent: "Conflict, embedded in a spirit of stubborn generosity, is not only generative but necessary because it acknowledges the undeniable—the social and economic substructures of power, of racism, of identity that will not be erased by goodwill" (51). From such conflict can emerge "a joint inquiry into thorny problems, opening up live options that let us construct a language of possibility and a more complicated ground for action" (50-52). It is important for students to anticipate in advance that dissent and conflict will arise and to be ready to respond to it productively rather than wasting time trying to suppress, reform, or eject dissenters. Two textual presentations of dissent that my students have found useful are counterevidence and minority opinions. The presentation of counterevidence draws on established models of academic persuasion, in which a thesis (the opinion of the majority) is advanced but in which counterevidence as well as evidence is presented. In employing this option, students must avoid the approach to counterevidence that traditional argument offers. Counterevidence must not be discounted or "refuted"; rather, it should serve to enrich the thesis, showing its complexities and ambiguities. The presentation of minority opinions draws on Supreme Court practice. Collaborative writing groups employing this tactic present a final paper, to which is attached one or more statements of dissenting opinion. Nor are these statements individually authored; the entire group works dialogically not only on the majority paper but also on the minority opinion(s).
9. Explain in advance how the project will be graded, preferably involving the students in the decision. I told my sophomore composition class that each collaborative group would receive a single grade but that the groups would decide in advance how a shirker would be graded. I provided my criteria for grading, telling them that I expected a better product than I would of a single individual but that I expected sophomore- and not professional-level work. Some teachers try to assign individual grades for a collaborative project—a method that I cannot recommend, since it undermines the purposes of collaboration. Such quandaries can arise from the gatekeeping responsibility for judging and ranking individuals that is endemic to many writing programs (see Holdstein). Before assigning collaborative writing projects, the teacher should ascertain that the institutional purposes for the course and the teacher's own purposes in assigning collaboration are sufficiently harmonious that the institutional agenda will not undermine the collaborative pedagogy—or vice versa.

Sources