Composition was a craft once shrouded in secrecy. But ever since rhetoric and composition re-emerged as a field of discourse, many scholars have explored writing as a process. Early research dealt with “lengthy, idiosyncratic narratives” that were difficult to parse and draw conclusions from (Perl 194). Furthermore, they depended on the writer’s ability to accurately remember the reasoning behind certain decisions, which was not always reliable. In 1979, Sondra Perl expanded on this research in her study of unskilled writers and invention of a protocol coding system. Her work attempted “to elucidate basic processes” involved in composing and to shed light on how to improve the skills of poor writers (Perl 193).
Two years later, Flower and Hayes endeavored to unpack the processes they saw at work in writing protocols. Unlike the earlier stage-process model, theirs involved a hierarchical set of writing practices with fluid transitions between processes. Carol Berkenkotter’s work reflected this model while exploring how skilled writers compose in their natural setting. She argued for paying attention to “the context in which revision occurs,” as most protocol studies relied on a laboratory setting (Berkenkotter 218). Berkenkotter allowed her subject, Donald Murray, to work on his own writing rather than creating timed writing assignments. In so doing, she urged scholars to rethink their expectations regarding the sub-processes involved in composition.
To understand my own process, I conducted the following self-study; it helped me better understand the ways I synthesize information and approach the writing task. The results allowed me to compare the techniques I share with other writers and the ones that are unique to me. As many of the aforementioned scholars conclude, there is no “right” way to write. But studying the intricacies of one’s writing process can help one understand the circumstances that lead to good (or bad) writing.
To collect the data for my literacy self-study, I used QuickTime’s screen capture feature to record myself as I began to draft my literacy narrative. I narrated all my thoughts and decisions as I began to make sense of the prompt, the readings, and my perspective. Perl and Berkenkotter used a similar methodology, which they termed a “thinking-aloud protocol” (Berkenkotter 219).
Two weeks later, I returned to my recording. This time, I used Perl’s coding system and a timeline sheet to note down each of my writing, reading, and editing behaviors as I composed my narrative. According to Perl, the coding system “provides a way of analyzing the process that is categorical and capable of replication”—indeed it does, since I have replicated it here (Perl 197).
My coded sheet revealed the extensive amount of time I spend planning before committing myself to original prose. In fact, I wrote barely more than four sentences by the end of the 40 minutes. During the planning stages, I re-read the prompt over and over until I ultimately pasted it into my Word document for easy access. Because the prompt was broad, I read aloud passages from the readings and jotted down the ones I thought might help me narrow my focus. Even once I began to draft the essay, I quickly returned to planning to re-evaluate the direction of the essay and consider other possible routes.
Another recurring code on my sheet was Q, which stands for asking a question. I asked questions aloud about what particular ideas from the readings meant, such as, “But does she think there’s nothing positive about the current-traditional method?” As I framed the idea of literacy in my own academic background, I asked myself questions about how certain experiences related to the readings and the prompt. For example, I wondered aloud, “Learning to write and rewrite was a big part of my development as a writer…but I feel like literacy is more than just writing, isn’t it?”
Because I spent little time typing words, I spoke most of my thoughts out loud. I made comments about the ideas expressed in the readings, marked with C’s and I’s on my coding sheet, and reflected on their relevance to my own experience with literacy. I often said the phrase, “I want to say that…” before attempting to verbalize my ideas, especially once I began to draft the essay. Whenever I ran out of things to say, I repeated key points until new ideas dawned on me.
Although the study occurred in a natural setting (at my desk at home) and the writing assignment was organic (an essay for a class), I still felt watched. I never speak out loud while planning an essay, so I had to make a conscious effort to speak as the thoughts rolled in my head. I felt unnerved by long bouts of silence and consciously resisted the urge to check my Twitter. It felt somewhat like a performance—and I have stage fright. I think exposing my writing process required something akin to vulnerability, which surprised me, since I was the only audience member. And the timed aspect made me feel “inadequate,” as Murray put it; I felt pressured to have something to show for 45 minutes of writing (Murray 231).
The numerous PL’s, or general planning, made me realize how much time I spend planning before writing. Unlike Perl’s subjects, who often “began writing without any secure sense of where they were heading,” I dislike typing anything before I gather my thoughts (Perl 205). Once I’ve typed something, I feel as if my options have narrowed. Although I am aware of the fact that I can delete anything I’ve written, each sentence I write feels like a commitment. Flower and Hayes describe this phenomenon as one of the forces in competition during the act of composition, as “each word in the growing text determines and limits the choices of what can come next” (Flower and Hayes, 371).
For this reason, I wrote little and thought (or rather, spoke) much in my recorded session. My frequent use of the phrase “I want to say that…” can be traced back to this impulse: by verbally explaining my ideas first, I can be sure of my intended meaning before committing to it. I studied and commented on the readings before fleshing out my own thoughts so that I wouldn’t be limited by dead-end ideas. In a normal setting, I usually write my nascent ideas on paper, which feels less confining than typed words.
The behavior that surprised me most, though, was one that did not have a corresponding code in Perl’s guide. Several times during my planning stages, as I typed up a few bullet points describing my main ideas, I swiped left on my touchpad to access the thesaurus on my dashboard. In these moments, I wasn’t trying to find the most polished word or clever phrasing. Rather, I used the thesaurus to further jog my ideas about the topic. For example, I entered the word “literacy” into the thesaurus early on in the recording; the results included “ability to read and write,” “reading/writing proficiency,” and “book learning.” I said out loud, “Hmm…I definitely did a lot of book learning growing up,” which then prompted me to consider how the current-traditional approach had influenced my writing style. In this way, I worked similarly to the unskilled writers Perl studied, who often began an assignment by forming “a string of associations to a word in the topic” (Perl 205).
This self-study helped shed light on the idiosyncrasies of my composing process. I noticed the behaviors that had “become more automatic than deliberate,” such as the planning and thesaurus-searching (Berkenkotter 228). I now understand the reasoning behind certain writing rituals I have, such as drafting on paper before committing to type. Hopefully, these results will allow me to better recreate the conditions and processes that lead to good writing.