Professional Learning Teaching Writing

Elbow Room: Tweaking Response in the Secondary Classroom


One Tuesday morning three years ago, in an act of semidesperation, I changed the way I teach writing forever. Frustrated with the lifeless prose from my eleventh grade class and thrashing around for a way to combat my students’ ennui, I pulled out of my briefcase a messy cover letter for a summer job that I had begun to draft. I’d put off revising it and had thought of abandoning it altogether. But I reasoned that if I showed my students this draft as something I’d worked on and struggled with, they could get a sense of some real-world writing in process. It would be embarrassing, of course, to reveal my awkward prose, but at the moment, I was willing to try anything. What did I have to lose? Maybe they could even help me revise it. I shared the letter with them. I was not prepared for what happened next.

They didn’t just respond. They stood up and yelled. Disaffected students who hadn’t participated in class in days were eager to tell me what they thought.

“Your opening is so weak! I wouldn’t hire you,” someone offered.

“That one line’s funny. Keep it,” from a boy who swore he wasn’t smart.

From another, “That part about when you were working at the Gap, that’s good. I get it.”

A girl with a well-known contempt for school added, “Your second paragraph…I don’t know. It isn’t as good as the other ones.”

I was astonished. We were talking about writing, and they were liking it. What’s more, while the students’ comments were pretty general in a way, they were also right on target. My opening truly was lifeless, but the Gap story did work pretty well. I left that day, wondering if my students secretly knew more about good writing than I thought they did. How could I direct these burgeoning skills and enthusiasm toward their own writing?

The events in this brief narrative caught me at a time when I was asking myself a lot of questions about my teaching. After a few years of happily teaching English, I found myself wondering why some of my classroom instruction worked, and why some failed (see Blau 1988, 32–33). One year earlier, I’d participated in a summer institute at the Pennsylvania Writing and Literature Project (PAWLP), which presented me with so much pedagogical richness I’m still unpacking it. A theory that struck a particular chord with me was Peter Elbow’s ideas for responding to writing, featured in his book Writing Without Teachers (1998b) and other texts. He advances the notion that the main function of response should not be peer editing, but rather audience reaction. This idea seemed to speak to the real-world writing strengths and needs of my students. It took a year and a half of incubation, and that breakthrough day with my juniors, for me to implement the response groups in my class. I haven’t looked back.

Though Elbow normally works with students and writers older than the high school students I usually teach, I believed I could adapt his ideas to make them viable in my classrooms. But this would take some research, some trial and error, and close observation of my students.1 I later found Denny Wolfe and M. Lee Manning’s articulation of this process: “Often, classroom-based research demands nothing more than a) being more attentive to what students are doing and how they’re doing it; b) recording our observations; c) trying to make sense of our recordings; and d) making adjustments” (1997). With only slight jocularity I might add to that list “willingness and patience,” and “respect and appreciation for the students you involve in this endeavor!” This article is in part a review of Elbow’s methods, and in part about the “adjustments” I made so that his ideas would work in my particular contexts.

Beginning Elbow

Here is Elbow’s basic response-group structure. Over a semester2, writers share a piece of writing they wish to improve with a group of six or more of their peers. The piece may be new or it may be one the writer is revising. Each author brings in enough copies of the work-in-progress for everyone in the group. He or she reads it to the group at least twice. During the readings, the audience members write their reactions to the piece. When everyone has had a chance to react in writing, the group begins to respond aloud to the author, who listens and takes notes. When one piece is finished, a process that takes about 15–20 minutes, the group moves on to the next author and the next round of responses. While Elbow’s writers focus on a single piece, in my classes, each writer produces five or more pieces, which circulate through the groups as drafts and later as revisions.

To let students take full advantage of the revision process the groups foster, I had to rethink the way I structured writing assignments. In the classes I discuss here, students were given a variety of writing prompts (primarily creative nonfiction) and asked to develop between four and six of them into full-length works of 1,200–2,000 words. I let the students know in the beginning of the course they would need multiple copies of their works. Many students were comfortable composing on computers and printing multiple copies. I also gave time to photocopy handwritten drafts in the library.

Though the majority of my teaching schedule for the past few years has been homogeneously grouped junior-level English classes, I also teach creative writing classes, college composition classes, and a summer workshop for middle school students. I have run response groups in all of these environments. I summarize the classes in Table 1. For the purposes of this article, I focus on my experience over three years with high school juniors in two courses: a mid-level class with a rhetorical emphasis, and Advanced Placement Language and Composition. There are no placement criteria for English courses in our district; student choice determines enrollment. The differences between the groups’ performance, though interesting, are beyond the scope of this article. I focus instead on what worked in both situations.


Like many English teachers, I’d tried peer review activities before. What makes Elbow’s response groups more effective than some other models for response is the kind of response the audience gives. Elbow’s concern is with what worked about a particular piece of writing at a particular time in a particular reader’s head. Students give detailed retellings of what they thought, remembered, saw, or did as they read the words. With enough of these responses, writers start to realize the effect their prose has on others, what kind of echo their words make when they throw them out to the universe. The response groups give writers a real live audience who tell the truth about what they liked and why.

This type of response levels the playing field. Some students say they don’t feel qualified to give a response; they are afraid of giving the wrong advice. The response group’s focus on meaningful reaction, not grammar-level editing, helps allay these kinds of concerns. Eventually, with practice, comments like this one from Caitlyn surface: “I felt qualified. I knew that I wasn’t the best writer in the world, but I knew what I didn’t understand, and what wasn’t entertaining, and that it was my own opinion. So I felt qualified to say it.”

When students understand these rules of response, they also avoid the trap of responding with “that’s good” or “I like that.” They don’t focus on grammar, spelling, or usage errors in place of content and voice. (I’ve even heard them playfully correct each other when one slides into those less-helpful comments!) One of the values of personal responses is that authors are free to take them as opinion, not command. It’s their piece, I remind them. They’ve got to be happy with it. Kelly remarked on the artistic freedom the groups offer: “We didn’t have to take people’s advice. What everyone said were just their opinions, not set in stone. We knew taking people’s advice was an option.”

Of course, preparing students to give and take this type of response requires some training. During one of my earliest response groups, eager to capture a kind of transcript of what was happening, I carried my laptop around to the groups and typed comments I overheard—the good, the bad, the funny. At first I used these as a review of correct response behavior. This list has since grown into a full-fledged PowerPoint presentation that is my introduction to the groups. It’s a hybrid of Elbow’s theory, my advice, and my students’ comments. I offer a few sample slides in Figure 1.

Critics: do this!

  • Bottom line: let this author know what got through to you!
  • Write as the author reads, as you read, and as others respond. Hand
    these to the author.
  • Make sure you’ve had a good chance to read the writing.
  • Give specific reactions to specific parts—passages, lines,
    phrases, individual words.
  • Try to use statements beginning with “I.”

What to say: quotes from students

  • I got the picture
  • this part made me think . . .
  • these lines are effective because . . .
  • that sentence makes me believe what you’re saying
  • I saw this . . . I was right there with you
  • I think you got the point across that you . . .
  • that part was key to showing your effect
  • I was confused by . . .
  • I’m having trouble with . . .
  • it tripped me up a little . . .
  • I was lost at . . .
  • this broke the flow for me . . .
  • I’m still not sure what you’re trying to say . . .
  • I don’t quite understand why . . .
  • I think you were stronger in your first paragraph
Figure 1: Sample slides from group training

I’ve also added to Elbow’s structure a brief but useful modeling activity. Before we begin the groups formally, the class looks at something I’m working on. (I’ll never forget that first time!).3 I read it aloud; they read it silently. Then I encourage everyone to give a response, pulling from the charts if they wish. I let them see how the “I” statements work. (“I was with you there; I saw exactly what you were talking about.”) The statements keep editors telling the truth and authors from feeling like they’re being attacked. I model the appropriate nondefensive author behavior that includes taking notes and responding gratefully to helpful comments. I listen, nod, and say little but, “Thanks.” This is great fun for me, prodding them to take my writing to task. Of course, my students have helped me immensely with this article. They’ve seen it through many drafts, I’ve shown them the valuable revision suggestions from The Quarterly editors, and they’ve said they will be excited to see this published copy. Modeling an activity is a potent device—we all know this—but it seems to gather a particular strength when, in front of your students, you practice what you teach.

Ultimately, however, there’s only so much pregroup preparation I can do before I hit the law of diminishing returns. I can sense when they’re itchy to begin, and I’ve said enough. They’re eager to get on with it, and the best training is the doing. My student David realized this: “This isn’t anything that can be taught. You can get the gist, but you have to experience it.”

The Time Factor

Elbow suggests fifteen to twenty minutes per piece for reading and responses (1998b, 84). Gradually, I have found twenty minutes per student to be a good guideline. That is, a piece is read twice—this takes about five minutes—and another fifteen minutes are given over to responses of about three minutes each. This gives enough time for meaningful response, but not so much time that students wander off.

Once, for a few weeks, I tried letting the groups run without time constraints. Students became bored, frustrated, and impatient. Off-task behavior occurred; efficiency decreased and dead horses were beaten. Besides, kids like to talk. “Sometimes the groups got off-topic, when we weren’t prepared or were bored with the same piece. We do need boundaries. We’re in high school,” remarks Alison.

Elbow is specific about how frequently response groups should meet: once a week, for at least ten weeks (1998b, 84). The ten-week commitment was something I could do, but in the beginning I was nervous about group sessions once a week. As we know, there’s other curriculum to teach. Honestly, I felt uncomfortable with what I considered “giving up” time from my curricular concerns, so the groups met infrequently. Students started asking for more time, so I took a hard look at what was happening. I had to acknowledge that the response groups were on task and discussing writing in meaningful terms even though they were meeting only about once every three weeks. But students had impressed me with the quality of their critiques, with how serious they were. So I agreed the groups could meet more frequently, and I haven’t looked back. Now, in all my classes, the response groups meet once a week. For organization’s sake, I earmark the same day every week for these groups (see Elbow 1998b, 84). When Wednesday comes, for instance, we all know what we’re doing.

Forming Groups and Rotating Group Members

Elbow also recommends that each group have no less than six members. Though I was wary of making groups larger than four, I have come to agree with Elbow that larger groups are appropriate. The premise behind the response groups is to create the experience of voices in an audience, a polyphony, a varied sample. Different people have different reactions. It is the multiplicity of reactions that helps authors realize the effect their prose will have on any given audience. I’ve noticed that the larger groups (I’ve gone to groups as large as eight or nine) provide a kind of formality, some emotional distance from the personal material. It’s a vulnerable situation, to be sure, sharing one’s work with peers. The larger number of participants helps keep things from getting personal. There’s no whispering, no cliques. The group is too big for that.4 And it’s a sizeable audience, too, so there’s some pressure to perform responses well. When I ran a brief version of the groups with middle schoolers, even they seemed to sense the formality in the informality, and stayed focused on their task without being watched over by me. Everyone was eager to be read and heard.

But in a serious departure from Elbow’s original response group model,5 I do not keep group members together for the entire semester. This, too, is the result of much trial and error. When I tried keeping groups together, there were problems. The students got bored with the sameness and reported that peers’ responses became predictable and, in that way, less helpful. As Olivia observed, “The thing is, as they start to get how you write, they start to miss the things you miss, they’re turning into you, so things get missed more that way. They recognize things the way you would see them and not like an audience.” Kelsey agreed: “When group members look at the same piece more than three times, it starts to lose its effectiveness due to the boredom of reading the same material.”

Since there are as many as four groups running simultaneously in my class, I began by suggesting that the groups change members every other meeting. This was a disaster. Chaos ensued, and we lost too much time in the shuffling. Then I tried mixing the groups once a month. This met with more success, but still wasn’t quite right. Finally my students made a suggestion, which I use to this day. Each group member has a number. Each meeting, one or two numbers rotate. For instance, I’ll say “Today, all number threes move one group to the left,” or “this week, number sixes and twos move to the right,” or whatever mix I feel will benefit the class that day. This arrangement offers both consistency and freshness. It also allows me a surreptitious way of separating students who are not able to work well together. Sarah commented, “I liked switching groups; I liked the fresh outlook on what I wrote. Your group gets used to your style, and that’s nice, but I liked having other people looking at it.”

The Teacher in the Teacherless Class

One day while the groups were meeting, an administrator stopped in to speak with me. I had been giving her my full attention for a few minutes, when she suddenly stopped talking and looked around the room.

“They’re not looking at you,” she whispered.

“Right,” I replied, a little confused.

“They’re not even watching for you to be watching them,” she said conspiratorially.

“Ah, no, they’re not. They don’t really need me right now.”

I realized what was happening, and I smiled. She left, saying she had to tell other people about these “genius” groups.

A teacherless writing class, truly. But while they’re working, what do I do? Plenty! I monitor the groups, especially the first few sessions, to make sure that things go smoothly and that students aren’t slipping into less useful response habits. But the more completely I prepare them, the less they need me. Then I’ve got to let go. This is a teacherless writing class. Elbow suggests teachers join the groups as members, bringing their own writing for discussion. I’ve done this on a few occasions, and it’s been great fun. I abide by my own rules, my students give excellent responses, and they tell me they enjoy my being a member of the group. But as fun and helpful as it’s been, I don’t do it consistently, simply because I can’t pass up this opportunity for individual conferencing.

The response groups give me the precious commodity of time when all eyes aren’t on me. After the second session, or when I’m sure they don’t need me any more, I set up conferences. This I love, as do my students. Students sign up to meet with me for ten minutes during the period. While the groups motor along, I conference. Though the content and structure of these conferences is another topic, the students look forward to this one-on-one time with the teacher. The student knows he has my (nearly) undivided attention, and can use the time as he most needs.

I encourage students to bring to me their writing, with any particularly pesky areas. My responses to the writing differ from the groups’. I’m more likely to give examples or models the students might want for inspiration. I am (as English teachers are expected to be) what Elbow calls an expert reader. I’m likely to see things in the students’ writing they didn’t know were there. When someone wants to talk about her writing, I follow the same guidelines as the groups. Megan saw the groups and conferences balancing each other. “Your comments filled out what the peer groups had missed. Overall, the response groups and the teacher/individual conferencing worked very well together.”

Grading and Accountability

Elbow’s response groups, as he envisioned them, are voluntary meetings of writers coming together outside the school setting to improve their craft. Clearly, this is not the case in my groups. I’m changing the game by making response part of my course. Participation is required. While participation is its own reward, it’s also part of a grade. I have experimented with different strategies for giving credit or points for the revisions, for participation in the groups, and of course for the final products. I continue to experiment, weighing things slightly differently each semester, but I find it’s good to keep in mind the different skills at work here and reward each of them.

A word about paper grading load: if you’ve been following along, you’ve learned that students show each other a piece of writing regularly. Either the piece they bring to the group is entirely new or it has undergone substantial revision. That means my students are writing and revising constantly but are looking to each other, and to themselves, for evaluation and revision suggestions. That is, most of the time, they’re not looking to me for what to do, for what’s right. And I reinforce this independence and interdependence: I remind them that in a few short months, they won’t see me again, so spending too much energy wondering what I (or any other particular teacher) want is largely misguided. They must take responsibility, and soon they realize I’m not kidding. Our process includes only two times when I read student work: once when I’m asked to (perhaps in a one-on-one conference), and once when the student decides it’s done. In conferences, my responses are verbal and written, talking chiefly about issues the student raises. When the piece is handed in, it’s a final copy; it’s been seen by many sets of eyes and revised by the author. So I’m evaluating the final product. And then the comments are lighter because I’m not concerned with those time-consuming revision suggestions.

There Are Other Things Going On Besides Writing

Gradually, as I allowed the groups to take more time, I began to realize my students were doing more than discussing the specific elements of their own papers. They were becoming more conversant about elements of style. They were gaining control of their prose and generally building confidence in their ability to write and speak. The specific and careful ways the response group trains them to talk about each other’s writing also aid careful analysis of professional prose. And this isn’t just my wishful thinking: “When we comment on other peoples’ pieces, it made us more confident to comment on more famous author’s writings,” said Dani.

David, her classmate, agreed: “Not only did my classmates aid me in looking objectively at my own writing; the experience taught me how to look at others’ writing, be it a peer or a classic, with more objectivity and focus. The groups helped develop my analytical skills, which I have found useful not only in looking at other writing, but also in debate.”

I’m not suggesting the groups are a panacea. There are battles of ego; some students come unprepared, and tempers sometimes flare.6 But it can’t be bad to let high school students get to know each other. And they do get to know each other quite well, over time, not just through the interactions of the groups, but through the pieces they bring to the group. They hear a multitude of voices and are exposed to different viewpoints. They talk about issues raised in the writing that might not otherwise get class time—issues like college pressure, teenage motherhood, family dynamics. Practice in conflict resolution and respectful listening are valuable side effects of an activity designed to improve writing. Sasha remarked, “I think the whole entire process made us grow as people in our lives . . . you can relate to others, and they to you. It helped you see yourself and others in a different light. It helped me grow as a person and not just as a writer.” I could not have anticipated such outcomes—and they’re precisely why I continue the groups.

I offer as a closing some observations on trouble spots in my groups, and what we did to overcome them.

Troubleshooting #1: Anxiety. “We’re all really sensitive. We need praise to balance out the criticism,” observed Megan, a talented (and comparatively confident) writer. Many students agree with Megan’s sentiment, reporting a feeling akin to horror when they realize they will be reading their work aloud. Most of us have a sense of protectiveness, or defensiveness, about our creations, and students are no different. Elbow (1998b, 96) suggests avoiding negative feedback entirely for the first three or four meetings. By that time, the authors are itching to get criticism. That’s not to say everyone is always gentle; I still find some discreet monitoring is necessary to make sure feedback is positive enough to be helpful.

Careful instruction in the beginning about audience response helps decrease students’ fear of sharing their work. The group needs to understand that everyone is similarly terrified and yet similarly sharing. As the newness wears off, the anxiety wears off, too. Students begin to see this as an orderly, efficient group, not a gang of marauders. Lauren says, “I think I built a lot of confidence having to go into a group and read my writing. I never really appreciated my writing so much! You’re out there and exposed.” Lauren seems to have taken the enviable step to admiring her own writing from the outside.

Troubleshooting #2: Being cruel, only to be kind. “The response groups forced me to give up my protective nature toward my pieces. Once I allowed my group to truly criticize my work, I realized how helpful they actually were,” said another student.

It occasionally happens that responses are thin, or fluffy, or uncommitted—even when students are responding according to the Elbow model. When this occurs, I’m careful to look into the situation. If it’s early in the group process, the reluctance to respond may be because of the newness of the experience. This is easy to remedy: stick to the groups, and reluctance basically goes away with practice.

It’s possible, too, that students aren’t responding because they genuinely don’t like a piece and can’t find anything nice to say. This is a different issue. They are afraid to be negative. They don’t want to hurt anyone’s feelings.7 So, before the first group in which negative reactions are allowed, I give explicit permission to be honest and straightforward. I address the class about growing pains. Real growth is uncomfortable, I remind them, and writing is no different. They’re doing their peers a disservice if they say only pleasant things after the first few groups. Paraphrasing Elbow, I ask them to imagine how crummy it would feel if no one had anything helpful to say about something they’d worked on for hours. I also advise they use some of Elbow’s other techniques: summarize the piece, look for its center of gravity, or tell what they feel it is about (Elbow and Belanoff 2000, 511–513).

I’ve had success with another technique for shaking up an overly conciliatory group. I ask any willing author to state, before her reading, that she wants the group to be particularly brutal. In effect, she gives the other students permission to be extra tough. This is even more effective if she states areas of difficulty: perhaps she’s disappointed in a sentence or her word choice overall; perhaps it’s more a global concern over theme or imagery. Without this kind of permission, with specific reasons, groups can flounder, because darn it, everyone’s too nice.8


My findings here have been based on experience with dozens of different groups of students, ranging from fifth grade to adult returning students, class sizes ranging from nine to thirth-one with the most common in the high twenties (see table 1). The student statements are almost exclusively from junior advanced placement students, only because in these classes I took the most time for reflection and discussion about the group process itself (an important step which I’ve not been resolute enough in encouraging). As articulate and observant as these comments are, they may be exceptional in tone but are representational in content. Similar sentiments have come—and continue to come, even in the days I write this—from juniors and seniors of mixed ability levels, from community college students, even from my colleagues.

We know that the success of any technique, however well conceived, depends greatly on context. I believe, as do some of my colleagues who have adopted Elbow-style response groups, that these groups work in part because they call on skills that every person already has or can develop, most particularly the ability to articulate as an audience member a reaction to the parts or the entirety of a piece of work. We as teachers give the occasion a dose of our professional knowledge, and a helping hand.





Belanoff, P., M. Dickson, S. I. Fontaine, and C. Moran, eds. 2001. Writing with Elbow. Logan: Utah State University Press.

Bennett, C. 1997. “Teacher-Researchers: All Dressed Up and No Place to Go?” in Wolfe and Manning “Taking Charge of School Reform: English Teachers as Leaders.” The English Journal 86 (5): 36-38.

Blau, S. 1988. “Teacher Development and the Revolution in Teaching.” The English Journal 77 (4): 30–35.

Cochran-Smith, M., and S.L. Lytle. 1999. “Relationships of Knowledge and Practice: Teacher Learning in Communities.” Review of Research in Education 24: 249–305.

Elbow, P. 1968. “A Method for Teaching Writing.” College English 30 (2): 115–125.

Elbow, P. 1969. “On Peter Elbow’s `A Method for Teaching Writing’: Reply.” College English 30 (7): 591–595.

Elbow, P. 1998a. Writing With Power: Techniques for Mastering the Writing Process. 2nd ed. New York: Oxford University Press.

Elbow, P. 1998b. Writing Without Teachers. New York: Oxford University Press.

Elbow, P., and P. Belanoff. 2000. A Community of Writers: A Workshop Course in Writing. 3rd ed. New York: McGraw Hill

Elbow, P., and P. Belanoff. 2003. Being a Writer: A Community of Writers Revisited. New York: McGraw Hill.

Haskell, D. 1991. “EJ Update: Show Off & Tell: Another Look at Teachers Writing with Their Students.” The English Journal 80 (4): 65.

Queenan, M. 1987. “Classroom Inquiry: Teachers as Researchers?” The English Journal 76 (4): 88–90.

Wolfe, D., and M.L. Manning. 1997. “Taking Charge of School Reform: English Teachers as Leaders.” The English Journal 86 (5): 36–38.


1. “Teachers need to re-define their roles to include responsibilities for conducting their own research with students they teach” (Bennett 1997). See also Blau 1988, 33–34.

2. Elbow (1998b) recommends at least 10 weeks.

3. “[T]eachers of writing must write . . . their authority as teachers of writing must be grounded on their own personal experience as writers—as persons who know first hand the struggles and satisfactions of the writer’s task” (Blau 1988, 31).

4. Elbow does acknowledge the usefulness of pairs working together for responses, and recommends using them in conjunction with the larger groups (Elbow and Belanoff 2003).

5. Later (Elbow and Belanoff 2000), Elbow addressed the need for “fresh faces” and gave suggestions for rotating group members.

6. See “What kind of people?” (1998b, 79).

7. Interestingly, in my experience younger students are less likely to feel this way than adult students.

8. Elbow discusses a related trap in “How to destroy the class secretly” (1998b, 114).