Skip to content

Composition Forum 29, Spring 2014

Research as a Recursive Process: Reconsidering The Composing Processes of Unskilled College Writers 35 Years Later

Bookmark and Share

Sondra Perl

Research, like writing, is recursive. We go back to what we know in order to figure out where we might want to go next. Or, said more simply, we go back in order to go forward. This is one of the basic findings about composing that emerged from my study of the composing processes of unskilled college writers that appeared in the December 1979 issue of Research in the Teaching of English under the same name.

Now, almost 40 years after I conducted this research, I have been invited to revisit that study and to think about the assumptions it was based on. I have been asked to consider what I might have done differently and what of value remains from the work and the approach to research I used. I have been asked, in essence, to go back in order to go forward.

The article I am discussing is a synopsis of my doctoral dissertation, Five Writers Writing, which I defended in 1978 at New York University. During my years as a doctoral student, I was also teaching full-time at Hostos Community College of the City University of New York. It was an exciting time to be a new instructor in a newly opened community college: Open Admissions had welcomed thousands of underprepared students into the university, and English faculty across CUNY were trying to understand the task that had fallen into our collective laps. The CUNY Association of Writing Supervisors (CAWS) had just formed and met monthly, often on Saturday mornings; Mina Shaughnessy was collecting and analyzing thousands of writing samples written by SEEK students at City College{1}; and no one seemed sure about the most effective method for teaching writing to this population.

In 1973, as part of my doctoral studies, I read Janet Emig’s groundbreaking study, The Composing Processes of Twelfth Graders. Her work struck me immediately as important, forward-looking, and relevant to our concerns at CUNY. It was obvious to me that those of us teaching writing on CUNY campuses had no understanding of how our students wrote. In other words, we had no knowledge of their composing processes—what they did or didn’t do while writing—and I was excited to think that this topic could be the basis for my doctoral research. I also soon realized that I had a ready and willing population: the beginning writers in my first-year writing course. At that time, there were no requirements for permissions, no IRBs to submit to, no institutional oversight other than that provided by my dissertation committee. I was free to design the study in any way I saw fit to help answer the questions I was asking.

I did, though, feel compelled to follow the best methodological advice around. It came from the 1963 text that those of us in the early days of composition held as our bible: Richard Braddock, Richard Lloyd-Jones and Lowell Schoer’s Research in Written Composition. Summarizing 504 research studies in order to report on “what is known and what is not known about the teaching and learning of composition…[in] a special scientifically based report” (1), the authors famously indicted almost all of them, claiming that the state of research in writing could be likened to research in the chemical sciences in the days of alchemy. The quote is worth citing in full:

Today’s research in composition, taken as a whole, may be compared to chemical research as it emerged from the period of alchemy: some terms are being defined usefully, a number of procedures are being refined, but the field as a whole is laced with dreams, prejudices, and makeshift operations. Not enough researchers are really informing themselves about the procedures and results of previous research before embarking on their own. Too few of them conduct pilot experiments and validate their measuring instruments before undertaking an investigation. Too many seem bent on obtaining an advanced degree or another publication than on making a genuine contribution to knowledge, and a fair measure of the blame goes to the faculty advisor or journal editor who permits or publishes such irresponsible work. And far too few of those who have conducted an initial piece of research follow it with further exploration or replicate the investigations of others. (5)

This chilling indictment gave doctoral students like myself pause. We wanted to be part of a movement in which writing research was valued, where our studies would be seen as rigorous, where our findings would count. To do so, we were led to believe, would require that we model our studies on the experimental sciences where, in order to make informed judgments and comparisons, variables were controlled. We were, in other words, advised to learn how to conduct research based on the experimental method employed in the natural sciences.

So, although following Janet Emig, I wanted to look at individual writers, creating case studies of each one, I also felt compelled to control as many variables as possible: I would, as a result, give each student the same topics; I would make sure each student was allotted the same amount of time in which to write; each session would take place in the same sound-proof room in the college library; I would assume a non-interfering presence; and I would ultimately write my dissertation and the articles that grew out of it in the third person, in a voice that was both distant and devoid of personality.

I would now say that at its best the study was a hybrid, combining close observation of individual writers at work with a certain set of controls that were imposed at the outset by me, the researcher. At its worst, one might argue that it embraces a set of contradictory postures: the study wants to observe a phenomenon as it unfolds (which happens best, I now think, in naturalistic settings), but it imposes conditions that turn the situation into a rather contrived affair.

Would I set up the research in the same way today? I think not. Since the early 1970s when I designed this study, I have come to think that experimental methods used in laboratories are not well suited to studies that take place in and around classrooms. I’ve come to think of classrooms and schools as distinct cultures more suited to the methods of cultural anthropology (i.e., involving fieldwork and participant-observation) than the methods of natural science. I’ve come to embrace a human science approach in which researchers enter settings (e.g., writing classrooms) in an attempt to understand how the people in those settings make sense of their lives and their work.

I understand what Braddock, Lloyd-Jones and Schoer were trying to do, and I sympathize with their goals: attempting to move what was a fledgling field into a research arena that depended on strict controls for its validity—something that continues to be the appropriate approach for research in laboratories in, for example, biological sciences and medical research. But advances in our understanding of human behavior and the models that have emerged to understand it have moved many of us in a different direction.

As I moved from my dissertation research to designing research that took place in classrooms (see Perl & Wilson, Through Teachers’ Eyes: Portraits of Writing Teachers at Work), I came to think that the experimental method was not a good fit for studying the complex business of teaching and learning. For one, the variables that could be controlled were too small to matter, and secondly, the areas I was interested in were too complicated to control. Basically, I knew that no one teacher is like any other, even those implementing the same pedagogical approach. And students, like teachers, are individuals with different motives, desires, knowledge, leanings, etc. This distinction alone argues for research methods that are sensitive to the fluctuating changes that occur in daily living. Finally, it became clear to me that the kinds of insights about teaching and learning that I wanted to arrive at—understandings that involved human beings making decisions about how and what they taught and learned—could best be documented and understood by coming to know the people and what their lives and learning meant to them.

I wanted, in other words, to understand the life of classrooms and of writers as they unfolded, as they were lived and experienced. I wanted to be immersed in the settings in which these events occurred. I wanted to use my own notes and responses to the setting as guides to a developing understanding. I wanted, above all, to focus on the meanings of events and actions as these emerged in the minds and the writing of the individual people who lived in these settings. None of this fit within the experimental models championed by Braddock, et al, that I had adapted for my dissertation.

Perhaps I can best explain my desire for a different approach by referring to the methodologist whose work greatly influenced mine: Max van Manen and his approach to what he calls “human science.” According to van Manen,

[t]he preferred method for natural science, since Galileo, has been detached observation, controlled experiment, and mathematical or quantitative measurement….In contrast, the preferred method for human science involves description, interpretation, and self-reflective or critical analysis. (4)

How like my first study is van Manen’s description of the natural science approach: so fully committed to control and detachment. And how much more comfortable I became when I realized that if I wanted to conduct research on teaching, learning, and writing, I could best do so by entering settings and describing what I saw in as honest, open and reflective a manner as possible. Such descriptions and understandings—what van Manen refers to as phenomenological inquiry into lived experiences—emerges from closeness, not detachment; from being a part of the context, not separate from it; from coming to know the participants in their natural settings and discovering what events and actions mean to them. Influencing me as well at that time was Elliot Mishler’s important article in the Harvard Educational Review, Meaning in Context: Is There Any Other Kind? His title said it all.

But this is not to say that The Composing Processes of Unskilled College Writers did not reveal important patterns or that the tools I created to analyze the composing processes of my students were not useful. Far from it. In fact, at that time, I would not have been satisfied with observing and describing the composing processes of one or several writers and leaving it at that. Given how limited our knowledge of composing was and how little we knew of unskilled writers, I wanted to see if their seemingly randomly written products, studded with misspellings, misuse of words, and grammatical and punctuation errors, were more than haphazard attempts to put ideas down on a page. I wanted to see if they had, in fact, any sort of composing process at all. To do so, I would need to see if I could detect patterns—behaviors that recurred consistently over time. And to detect patterns—a research goal I still adhere to even in the most naturalistic of studies—I needed to invent an analytic tool. When students and colleagues look at the “composing style sheets” I created—and the long list of abbreviations I used to code the movements made by each writer—they either gasp or groan. How, they often ask, could something so abstract, codified, and schematic actually depict the flow of composing? The style sheets are, I admit, somewhat off-putting. But creating them taught me something I stand by to this day. That is, that getting close to the data one has collected, immersing oneself in it, finding ways to categorize and code it, is an invaluable process. The specific codes are, at this remove, far less important than the act of creating them. To do so, I had to listen to each composing tape over and over again until I was certain that I was accurately describing what I was hearing. And this going over and going back, again and again—another recursive move—taught me what was going on when my students were composing. A computer program analyzing a chunk of data could not have done this. It took my ear and my judgment to ascertain what was happening. The style sheets may look cumbersome. But what they taught me could not have been learned in any other way. I submit, then, that whether one uses charts or graphs or other coding devices, there is no substitute for deeply immersing oneself in the data one has collected and sorting through it until regularities or common themes emerge.

This immersion and the style sheets enabled me, in fact, to see patterns within one student’s composing over time and then patterns across students. I could, using the style sheets, show not only that recursiveness in writing is an established pattern within the composing of unskilled college writers but also where and how this recursive behavior shows up: both in support of composing and in ways that derail it. It was really quite a heady experience to be able to document these behaviors and show how they unfolded over time.

Coding composing processes did not, however, have wide appeal. Several researchers went on to use the method in future studies (Lay; Raimes), and I occasionally asked my own students to tape themselves while composing aloud to see if they could chart their own composing processes. But I suspect that the idea of coding and the look of the actual style sheets were just not that inviting to members of our profession. What I find most fascinating now several decades later is that the tools of new media are enabling researchers to make use of all sorts of “distant reading” approaches (Moretti), from data mining (Lang and Baehr), to text mining (Mueller), to aggregating large amounts of hand-coded data into computer-generated graphs and charts (Miller).

I applaud these new methods and believe that they will enable us to understand and envision our field in entirely new ways. My only caveat is this: Human science requires the presence of an inquiring mind and a sensibility attuned to the nuances in the data. Only the researcher attuned to such complexities can make the leap from codes to charts, from charts to patterns, and from patterns to insights.

Given the controls I imposed on the student writers and the coding I created to analyze the data, readers of this early work might be tempted to conclude that I remained several steps removed from the students. Had the participants not been students from my own classroom, it is indeed likely that I would not have had a way to develop a relationship with them nor would I have been expected to. But that, happily, was not the case. Our classroom interactions enabled me to gain insight into my students beyond the confines of the study, which, I assert, helped me see them as the separate and distinct individuals they were and not merely as research subjects.

So, too, in subsequent research, my focus has remained tied to individual teachers and writers and to the ways they understand and make sense of their worlds. Van Manen would explain this as putting “persons” rather than categories first, as paying attention to the meaning of events as they unfold in idiosyncratic ways for individual learners and writers, as making sense of the singular.

To convey this sense, I found it important in subsequent research to cultivate a different kind of researcher’s voice, moving from third person to first, from a distant to a more personal stance and tone. It also became important to me to forego presenting my findings as a series of generalizable results. Instead, I began to tell stories. As early as the mid-1980s, when conducting and then publishing research on writing teachers in a K-12 public school on Long Island, I knew that storytelling would be my preferred method for writing up the results. While it would have been possible to structure the report of this research according to themes common to each class based on comparisons across the grade and school levels, what spoke to me—and to Nancy Wilson, my coauthor—was the singularly interesting and unique life experiences that unfolded in each classroom. We wanted, in other words, to preserve the integrity of each class we had lived in as participant-observers. We believed that our stories would show, better than any set of generalized rubrics, what life was like inside those rooms and what happened for students and teachers as they pursued a curriculum focused on reading and writing. In Through Teachers’ Eyes, we did include a final chapter that discussed themes, but the six main chapters of the book consist of six case studies, all told as stories—one for each teacher who welcomed us into his or her classroom for a full year.

One additional example may help to show how fully I now embrace what is also often referred to as narrative inquiry. In the 1990s, still at CUNY, having moved from Hostos Community College to Lehman College and then to the CUNY Graduate Center, I was invited to travel to Innsbruck to teach Austrian teachers of English who were enrolled in an MA Program at City College. I hadn’t planned to write a book about this experience; I originally thought I’d travel, teach, and return to New York. But the project began to call to me in the way that projects sometimes do: I could not stop thinking about the teachers; I was surprised by how much I wanted to work with them again. Most powerful of all, I realized that being in Austria raised all sorts of troubling questions for me, both personal and professional. The journey—or what I later began to think of as an odyssey—is described in what I call a teaching memoir, On Austrian Soil: Teaching Those I Was Taught to Hate. This book is, in my mind, an example of narrative inquiry written as creative nonfiction. In it, I use notes from my own journals, fieldnotes, and personal interviews, along with excerpts from teachers’ notes, journals, emails, and letters, to tell a story.

To compare this later writing to my early work in such articles as The Composing Processes of Unskilled College Writers is to see an entirely different voice and use of form. What was once distant is now close; what was once a clear and concise academic report is now a multi-genre personal account that reads more like a lyric essay. I hope On Austrian Soil exhibits what van Manen calls the hallmark of phenomenological inquiry as text:

…the cultivation of one’s being, from which the words begin to proliferate in haltingly issued groupings, then finally in a carefully written work, much less completed than interrupted, a blushing response to a call to say something worth saying, to actually say something… (8)

Van Manen can be seen here, perhaps without realizing it, as describing the recursive nature of composing: the halting nature of how words come to the page, the interruptions that, more often than not, return us to the page, sending us back in order to move forward, the dawning sense that there is always more to say.

As I began to study composing processes, what fascinated me and what continues to fascinate me today is where this “more to say” comes from. How do writers go from silence to words? From a budding awareness that they may have something to say to actually saying it? A question that prompted my early work and sustains the work I still do today asks how we know what to write next. Where in our bodies, I ask, does this knowing reside? The response to these questions is one that I began to articulate in my dissertation, that I hinted at in Composing Processes, and then developed more fully one year later in a now widely anthologized article called Understanding Composing, published first in College Composition and Communication. It was there that I introduced the idea of felt sense.

The concept of felt sense would, by its very nature, elude all of my attempts at coding. I could deduce that it was present, but only in the moments when students would pause (a codable behavior) and sit silently. As a researcher, I felt as if I had to account for what was happening during these moments of silence. I could not “prove” that students were contacting a felt sense, so in my dissertation and in the publications that followed it, I first raised this as a question: What was happening when students paused during the composing process? In the ensuing years, as I published about felt sense and gathered responses from writers of varying ages and levels of competence confirming the existence of a recognizable practice (pausing, sitting silently, and waiting for a felt sense of a topic to form), I began to feel more certain that what I had sensed or intuited in the composing behaviors of my students was actually occurring: namely, that in the silence, writers were going back not to words on the page or to topics they were pursuing so much as going back to an internal sense of meaning that they had yet to articulate either silently to themselves or visibly on the page.

It seemed to me, then, that if much of composing is recursive in nature, if many writers acknowledge that they go back (inside themselves) in order to go forward (constructing and developing their meaning on the page), then perhaps this way of composing could be taught, especially to students who have difficulty sitting still and listening to themselves, or to anyone who has trouble facing the blank page.

Much of my work in the intervening years in summer institutes at the New York City Writing Project, in my classroom at Lehman College, in seminars at the CUNY Graduate Center, and in workshops across the country developed from a desire to take what were originally findings from my research and turn them into effective ways to teach writing. The Guidelines for Composing that I created and were later published in a book called Felt Sense: Writing with the Body had their start, then, in my doctoral research. And that research had its start in the early, tumultuous, exciting, and confusing days of teaching writing at CUNY.

So while there are many things I would now do differently when designing research projects and writing up results, there are also many things that remain the same. In all of my teaching, writing, and research, I first and foremost endeavor to stay true to what matters to me: developing a writer’s voice that over the years has become increasingly personal and poetic; doing justice to the people I am writing about, attempting to show them as full, complex and caring human beings and including them, their voices and their responses as much as possible in the work I publish; and writing research that represents itself not as the final word but as a story—as my version of what happened. Looking back, I can see a clear line from Tony, the main case study in my dissertation, to Diane and Ross Burkhardt, two teachers featured in Through Teachers’ Eyes, to Margret Fessler, my dear Austrian friend and colleague in On Austrian Soil—all writers or teachers who have accompanied me as I learned the craft of research.


  1. As explained on the CUNY website, “SEEK stands for Search for Education, Elevation, and Knowledge. It is the higher education opportunity program at the senior (four year) CUNY colleges. It was established to provide comprehensive academic support to assist capable students who otherwise might not be able to attend college due to their educational and financial circumstances. Students are admitted without regard to age, sex, sexual orientation, race, disability, or creed.” See for more explanation. (Back to text.)

Works Cited

Braddock, Richard, Lloyd-Jones, Richard, & Schoer, Lowell. Research in the Teaching of English. Urbana, IL: National Council of Teachers of English, 1963. Print.

Emig, Janet. The Composing Processes of Twelfth Graders. Urbana, IL: National Council of Teachers of English, 1971. Print.

Lang, Susan, and Craig Baehr. Data Mining: A Hybrid Methodology for Complex and Dynamic Research. College Composition and Communication 64.1 (2012): 172–194. Print.

Lay, Nancy. Composing Processes of Adult ESL Learners: A Case Study. TESOL Quarterly 16 (1982): 204-214. Print.

Miller, Benjamin M. Mapping the Methods of Composition/Rhetoric Dissertations: A Landscape Plotted and Pieced. (Ph.D. dissertation in progress, Graduate Center, CUNY).

Mischler, Elliot. Meaning in Context: Is There Any Other Kind? Harvard Educational Review 49 (1979); 1-19. Print.

Moretti, Franco. Graphs, Maps, Trees. London: Verso, 2005. Print.

Mueller, Derek. Views from a Distance: A Nephological Model of the CCCC Chairs Addresses, 1977-2011. Kairos 16.2 (2012): n. pag. Web. 3 Oct. 2012. <>.

Perl, Sondra. The Composing Processes of Unskilled College Writers. Research in the Teaching of English 13.4 (1979): 317-336. Print.

---. Felt Sense: Writing with the Body. Portsmouth, NH: Boynton/Cook, 2004. Print.

---. On Austrian Soil: Teaching Those I Was Taught to Hate. Albany, NY: SUNY Press, 2005. Print.

Perl, Sondra and Nancy Wilson. Through Teachers’ Eyes: Portraits of Writing Teachers at Work. Portsmouth, NH: Heinemann, 1986. Reissued with a new preface, 1998. Print.

Raimes, Ann. What Unskilled ESL Students Do As They Write: A Classroom Study of Composing. TESOL Quarterly 19 (1985): 229-258. Print.

Van Manen, Max. Researching Lived Experience. Albany, NY: SUNY, 1990. Print.

Bookmark and Share

Return to Composition Forum 29 table of contents.