Essay writing (2 pages)

Grammar: Who Needs It? by Trygve Thoreson

Professor of English

William Rainey Harper College, Palatine, Illinois

A number of years ago my “Teaching English in the Secondary School” professor startled our class with the following pronouncement: “All the studies agree: grammar instruction does not improve student writing. If you really want your students to learn to write, you might as well just forget about teaching grammar.” The studies my professor was referring to, it turned out, had been brought to light in a comprehensive review of composition research published in 1963, Research in Written Composition by Richard Braddock, R. Lloyd-Jones, and L. Schoer, in which this statement famously appeared: The conclusion [of this review] can be stated in strong and unqualified terms: the teaching of formal grammar has a negligible, or because it usually displaces some instruction and practice in actual composition, even a harmful effect on the improvement of writing. The Braddock study’s findings received forceful confirmation some twenty-three years later in George Hillocks’ similarly named Research on Written Composition, in which Hillocks declared, “None of the studies reviewed . . . provides any support for teaching grammar as a means of improving composition skills.” Well. The evidence certainly seemed to be in, and it was “strong” and “unqualified.” Not a single study—not nowhere, not no how, it seemed—showed that grammar instruction had a positive effect on student writing.

This was thrilling news to me. I was determined to teach English somewhere, whether in high school or college, and in my mind “English” meant literature. I imagined myself whiling away the semesters bringing Twain, Hemingway, Frost, Dickinson, Dickens, and other standard authors to the masses, with (of course) term papers and personal responses and haiku and limericks and other creative and academically stimulating written exercises mixed in. In grade school and high school, I doggedly performed the usual Englishy tasks of the time—grammar worksheets, spelling quizzes, memorization of grammatical terminology (I can still see in my mind’s eye vividly colored posters with perky illustrations of the Imperative, Indicative, and Interrogative Moods), even some sentence diagrams—and I hated, or at least grudgingly tolerated, all of it. Grammar seemed to be the evil-tasting pill that had to be taken before getting to the literary dessert.

The idea that I might be able to spend an entire career as an English teacher without ever having to jam this particular pill down students’ throats came as welcome news. Following Braddock and Hillocks, in my early years as an English teacher I succeeded in keeping the grammar pill jar on the shelf. A teaching assistant at Northwestern’s Intro Studies program, I taught freshman English from a literary base, talking about thematic complexities and symbolism and literary theory while dutifully marking grammatical and usage errors on papers, but saying little about them in class. It seemed to work reasonably well, or so I told myself. The years rolled on, and with the years came new enthusiasms--“writing as process,” the social construction of knowledge, a variety of postmodernisms, peer editing, multiple drafting, portfolio assessment—yet in all this welter of innovation and change grammar instruction seemed to remain the unloved stepchild. In fact, it was practically viewed as the leprous outcast whom the villagers thought should be driven out of town. Grammar was something students intuitively knew, and it just had to be allowed to ooze out of them—naturally.

Here’s typical present-day advice on how students should make use of their “tacit knowledge of grammar” by the eminent composition specialist Peter Elbow:

Start off writing as naturally and comfortably as possible.

Don’t think about grammar or about any minor matters of phrasing or spelling.

Think only about what you want to say . . . .

Next . . . get your text to say exactly what you want it to say—but still without worrying about minor matters of phrasing, grammar, or spelling. . . .

Now turn your attention to phrasing, spelling, and grammar. . . .

[R]ead it aloud to yourself . . . and read your piece aloud to one or two listeners

Give your final, typed version to another person to copy-edit. [italics mine] (245)

In other words, forget about grammar until near the end of the process (these are “minor” matters, after all), and then rely on your own intuitive sense of what sounds right to make corrections. Then just hand it over to some poor sap who’s willing to do the boring work of tabulating scores. Such a deal!

For a time, I was perfectly willing to believe (hope?) this approach would work, but as the years went by, it became clear to me that it just didn’t. Student editors were just as clueless as student writers when it came to revising for correctness, partly because their knowledge of grammatical terminology and grammatical concepts was, shall we say, slight, and because in an environment where correctness is devalued (dismissed as “minor matters” or “mere surface error”) they had little motivation to expend much effort on such things. So I started teaching grammar. More precisely—perhaps, more outrageously--I started teaching copyediting (that thing that somebody else is supposed to do). I believed that certain conventions of Standard Edited English, such as where to put apostrophes, how to fix comma splices, what to do when subjects and verbs didn’t agree, could be taught and could be learned. I further believed that writers (including student writers) had to take the final responsibility for revising and editing their own work. If others could assist them and were willing to do so, well and good. If not, the ball remained squarely in the writer’s court.

I put together a series of tests—a diagnostic test to be taken at the beginning of the semester (with no effect on the student’s course grade), a second test over the same material to be taken at midterm, and a final test taken on the next-to-last class of the semester. Each test was followed by a review of each item and a discussion of the various rules/conventions that applied; little additional class time through the semester was taken up with questions of grammatical correctness. (Critics of grammar instruction are certainly right on at least one count: once you begin entering the thickets of grammatical information—and try to account for the myriad exceptions to every rule—the concern for correctness tends to take over, and the many other aspects of good writing don’t get adequate coverage.) Over time, I refined the tests, removing items that seemed trivial or that appeared infrequently in student writing, and including items I saw over and over again in student writing, whether some considered them “minor” or not. Thus, “different from/different than” was removed and “everyday/every day” included. “Who/whom” never made the test at all, since the distinction just didn’t seem to be a frequent issue in my students’ writing. I sensed that my instruction was making a difference, but I couldn’t be sure. I could absolutely document that students did better on the objective tests in the course of the semester (a receptive, motivated student could move from, say, a score of 17 out of 50 to a 43 or 44 in the final test), but it wasn’t entirely clear to me that these improvements were making a significant dent in the writing itself. And that, of course, was just the point that the authors of the Braddock and Hillocks studies had been making for years: that grammar instruction has no demonstrable positive effect on the actual task of composition (unless it is done only in the context of the student’s own writing).

So I spent a semester’s sabbatical leave assessing whether or not my tests made any difference in my students’ writing. In six sections of English 101 (four taught by me and two taught by my colleague Anthony Wisniewski), we compared essays written at the beginning of the semester (prior to any copyediting instruction) to essays done at the end (after the three copyediting tests and test reviews). A total of ninety students completed all of the tests and essays under the same conditions of test-taking, test-reviewing, and composing. Much of my sabbatical was spent simply counting errors in both sets of papers and then charting them to see how the objective test results correlated with error rates in actual student writing. Full results follow.

As will immediately become clear, I concentrated only on quantifiable error (and only on those errors Anthony and I covered in class); I made no effort to assess holistically the quality of the written work. This was partly because my own interest was more limited—I really just wanted an answer to a simple question: are my students reducing the number of errors (of the errors we’ve examined together) in their essays? It was also partly due to limitations in time and resources: to do a full-bore refutation of the Braddock and Hillocks research reviews (if I were so inclined), I’d need entire panels of outside readers, some very systematic and extensive controls on the project, including a control group of students who did not receive copyediting instruction, and so forth. Even if we had all these things, I confess to harboring a deep skepticism over how useful those results would be. I know from experience the complications and complexities of holistic writing assessment. Thus, for whatever it might be worth, here’s what I found. For my own overall conclusions concerning what these data suggest, please scroll to the just past the data grids. For an annotated listing of some of the secondary-source materials I consulted during this project, scroll to the very end.

Conclusions What does all this add up to? We can safely say that, in their written work, a clear majority of students (78%) showed some degree of improvement in their ability to edit out the errors covered by classroom instruction. About a quarter of all students (26%) showed significant improvement if by significant we mean a reduction in the error rate per 500 words of ten or more—an admittedly arbitrary definition of “significant.” One could argue that Student #15, whose objective score increased from 22 out of 50 to 41 out of 50 and whose essay error rate went from 10 per 500 words to 3.7 per 500 words, showed significant improvement. After all, having made just ten mistakes in the first essay, prior to any instruction, this student started from a position of strength and then was able to reduce his or her error rates to a level of near perfection, yet statistically we are counting this as only “modest” improvement. On the other hand, Student #67 improved “significantly” by reducing his/her error rate by more than thirty points in the final essay, but the student still committed 26.6 errors per 500 words, a high rate by almost any standard. Also not factored in is level of motivation for these particular assignments.

The first essay was diagnostic and thus was perceived by some students as a low-stakes project. The exit essay had consequences for the students' grades, and thus could be considered high-stakes. Overall motivation was examined, based on subjective judgments of the instructors. The boldfaced data shown above indicate students who were identified by the instructors to be exceptionally motivated students. (This derives from my hunch, from before this study was undertaken, that my most motivated students tended to benefit the most from this instruction.) The data, however, did not consistently bear this out. While 73% showed improved error rates on the final essay, a relatively high number of these were still in the “negligible” range, so that only 54% showed moderate or significant improvement, and fully 46% were in the “negligible” or “moderately worse” range. At least, none of them did significantly worse, which I suppose is a blessing.

So what does my sabbatical experiment prove? It may prove nothing. But it does, in my view, suggest something. Martin Luther King, Jr., is associated with a famous statement: “The arc of the moral universe is long,” he said of the civil-rights movement, “but it bends toward justice.” I’m well aware of my bold presumption in applying King’s elegant formulation to something as mundane as grammar/copyediting instruction. Yet an inelegant variant does express my impression of what the results say: in teaching student writing, the arc of grammar instruction may be wavy, but it bends toward correctness. Take that, Peter Elbow!

5