We Know What Works in Teaching Composition

Adam Niklewicz for The Chronicle Review

January 03, 2017

When I came to the University of Denver to start a campus writing program in 2006, I heard many faculty members say, "A lot of my students can’t even write a decent sentence." So when I read Joseph Teller making much the same assertion in an essay last fall, "Are We Teaching Composition All Wrong?" I recognized hyperbole when I saw it.

My response to that sort of exaggeration 10 years ago — joined by my 20 new colleagues in the writing program — was to gather and analyze a corpus of 500,000 words of student writing from classes across the campus. We found that, in fact, well over 90 percent of the sentences coded clear and error free.

Faculty members wanted to see better student writing (and I surely acknowledged and valued that desire), but it was clear that merely fixing sentences wasn’t going to achieve that end. There were larger issues: Students needed help developing and deploying their ideas and matching their writing with the expectations of various disciplines. Those things, we could work on.

Complaints about the state of student writing have a long lineage. In 1878, Adams Sherman Hill, a professor of rhetoric at Harvard, famously protested: "Everyone who has had much to do with the graduating classes of our best colleges has known men who could not write a letter describing their own commencements without making blunders which would disgrace a boy twelve years old." Hill and others devised pedagogies grounded in their own experiences and in common sense — though one man’s common sense was another man’s folly.

Teaching grounded in actual research took a scholarly turn in 1950, marked by the founding of the Conference on College Composition and Communication. Its journal is now the leading one in the field. By 1963, research on what worked in teaching writing — and what didn’t — had accumulated to a point that a synthesis was published, "Research in Written Composition."

Students needed help developing and deploying their ideas and matching their writing with the expectations of various disciplines.
Roughly 25 years later, George Hillocks conducted a new analysis (Research on Written Composition), using studies published in the intervening years. Since then, peer-reviewed research on the best ways to teach college writing has accumulated in dozens of books and well-established journals — including College Composition and Communication, Written Communication, College English, the Journal of Teaching Writing, Teaching English in the Two-Year College, Composition Studies, Writing Program Administration, and the Journal of Writing Assessment, to name but a few.

A 2005 article, "The Focus on Form vs. Content in Teaching Writing," analyzed why formalist approaches — like the back-to-basics kind that Professor Teller advocates — remained so popular in teaching composition, despite overwhelming empirical evidence that they were significantly less effective than other methods.

The teaching of writing happens — or should — within a deep field of practice, theory, and research. It’s also an enterprise marked by a fair amount of what Steve North, in a 1987 book, The Making of Knowledge in Composition, called teaching "lore." Lore consists of ideas and assumptions that are grounded in local experience ("what worked for me") and then passed along informally, for the most part, from one faculty member to the next. Lore is sometimes informed by research, and thus transmutable and generalizable, but more often it is not.

Teller’s essay participates in the tradition of lore. Not having been in his classes or having read his students’ work, I can’t judge his local experience, but I can judge how well his approach compares with the most effective national practices.

For example, his assertion, "Substantial revision doesn’t happen in our courses," might speak for his own classroom, but it surely doesn’t speak for mine or those of thousands of other professors. Consider his claim that students "do not use the basic argumentative structures they need." Again, while perhaps true of students in Teller’s own classes, that broad claim is unsubstantiated by my experience, by research on my campus, or by the wider literature in the field.

Where Teller departs most from actual scholarship in the discipline is his claim that "pedagogical orthodoxy" assumes that "composition courses must focus on product, not process." He could hardly be more wrong.

The two most dominant pedagogies today in college composition each focus on product as well as process. Genre approaches have students learn features that readers expect in specific kinds of writing (lab reports, op-eds, business proposals, magazine feature articles, movie reviews, and so on). Rhetorical approaches have students analyze the kinds of evidence, structure, and style that will be effective for particular purposes (for example, to persuade, inform, or entertain), for particular groups of readers (experts, novices, or people of particular viewpoints), and in particular situations. Both methods make significant use of model readings and examples.

One key to both approaches is sustained, guided practice. On that point, Teller and I surely agree. Students learn to write by writing, by getting advice and feedback on their writing, and then writing some more. What can be told to college students about writing can probably be encapsulated in a lecture of two or three hours. It parallels what meaningfully can be told about playing piano — the music notation, the relationship between notation and keyboard, the hand and finger placement, the posture, the pedal functions.

But without sustained practice on systematically more complex pieces ("Chopsticks" is not a Rachmaninoff concerto), the world’s best lectures will not — cannot — make a pianist. So, too, with writing.

Here is what this looks like in the best writing courses, informed by decades of research:

  • Students have ample opportunities to write. Professors expect them to write frequently and extensively, and we demand and reward serious effort.
  • Professors carefully sequence writing tasks. The idea is progressively to expand on students’ existing abilities and experiences.
  • Professors coach the process. We offer strategies and advice, encouragement and critique, formative and summative assessments.
  • Courses provide instruction and practice on all aspects of writing. Attend to the form and conventions of specific genres? Yes. Talk about creativity, invention (how to generate ideas), grammar, and style? Certainly, but also discuss things like logic and accuracy in writing, and how to fit a piece to various audience needs and expectations.
  • Courses use readings not only as context and source materials (which is vital in the academic and civic spheres) but also as models — and not only static models of form but also as maps to be decoded as to how their writers might have proceeded, why, and to what effect.
  • Professors teach key concepts about writing in order to help students consolidate and transfer skills from one writing occasion to the next. But we recognize that declarative knowledge is made significant only through practice and performance (see Bullet No. 1).
  • Student writing and student writers are the course’s focus. Everything else serves those ends.

Lore is a form of knowledge in every field. In pointing to the best practices of teaching writing — indicated by extensive research in composition studies — I don’t ignore the experience of individual teachers like Teller.

However, I can’t let pass unchallenged general claims about the way "we" are "wrongly" teaching composition, especially when they so dramatically misrepresent, even ignore, the field they would aspire to correct.

Doug Hesse is president of the National Council of Teachers of English and a professor and the executive director of writing at the University of Denver. He previously served as president of the Council of Writing Program Administrators and chair of the Conference on College Composition and Communication.