What Is This Thing Called Grammar?

(with some thoughts about teaching it)

The complex topic known as “grammar” causes enormous vexation in the field of teaching writing. I know of no one who can demonstrate that his or her mode of instruction will produce students guaranteed to write correct sentences. Not only that, correctness itself is the site of bitter disagreement within the profession. I will just note the following as a way of taking a position.

All children who do not suffer from some form of brain damage will learn to speak understandable sentences in the language of their parents. Grammar is hard-wired into human beings, from birth. In this sense, the vast majority of the human race is already grammatically correct, and the remainder are probably barred from achieving grammar by organic deficits which can’t be corrected by education. Though it’s obvious that language acquisition works, no one knows how it works. On top of that, I don’t believe that there is any consensus on how written language is acquired once the spoken language is in place.

Given that all of our students speak grammatical – i.e., understandable – sentences and can usually transcribe them with decent accuracy on paper, we have to ask ourselves what we really mean by “grammar.” If you examine a flawed piece of writing closely, you usually find that each sentence, considered in isolation, is a grammatical utterance, though it may be undesirably punctuated and words in it may be misspelled. The reader can understand each sentence even if it isn’t correctly copy-edited; for example, the most common punctuation error, the comma splice, usually does not interfere with the reader’s comprehension at all. The problem is in the thinking, the logic, the stringing of sentences together into a thought, the writer’s conception of what the key words mean – but not in the structure, the grammar, of sentences.

Beyond this, I think that under the category of “grammar” we are actually often talking about style or dialect. The real issue is often what we consider to be the style appropriate to an educated person in the context of college, or even more specifically, in the context of a certain discipline. Academic disciplines do have their own preferred styles, though each discipline tends to see its own style as neutral, invisible and obviously correct. I can’t emphasize this enough: people will generally see the style of their discipline as a universal standard, and this is simply not the case. Arguments about grammar and correctness are often misleadingly framed for this reason. “Why don’t you teach them to write correctly?” often really means “Why don’t you teach them to write like people in my suite of offices?”

Moreover, we are engaged in an institutional and national conversation about diversity which causes us to question the supremacy of cultural “norms.” This has direct application to the question of correctness in writing when students come from different dialect backgrounds, the obvious example being Black English Vernacular, which has its own internally consistent grammar that is “wrong” to the ear of an outsider. It is at least arguable that what we call “good grammar” is partly made up of signs of race and class background, and once that argument is made, it complicates the whole question further.

In short, we don’t know how grammar is acquired, we therefore don’t know how to teach it, we label as “grammar” some problems that are not grammatical, and we label as problems some features of writing that may be called cultural diversity.

Furthermore, if students don’t care about comma splices, it doesn’t matter how many times we teach them to use semi-colons instead; they won’t learn it until they care. And even if they care in a given writing class, this does not guarantee that they will care the following semester in some other classroom.


Here is a general outline of a strategy for addressing these issues in practice – though the outcome is not by any means guaranteed:

1. Don’t correct every mechanical error when commenting on student papers. It’s not our job to clean up our students’ writing. It takes too long, for many teachers it is too irritating and sours the relationship with the student, and it does very little good, if any. When the instructor corrects the student’s mistakes, the student generally feels “Oh, she’s taken care of that for me.” We are not here to write the paper for the student, in this way or in any other.

2. Therefore, circle or otherwise indicate the location of mechanical errors and move on.

3. Look for patterns of error in a given student’s work and, in your final comment, point out to the student one to three such patterns. (Why one to three? Because the student can only work on so many things at once. Why patterns? Because the point is to see the forest rather than the trees – to change the way the student thinks about constructing sentences.) When pointing out these patterns, try to show the student two things: the mistaken sentence logic she is apparently applying when she writes, and the more correct way to think about it.

4. Ask the student to apply the principles (the logic) outlined in your comment, not only in copy-editing the current paper, but in future ones.

5. Keep some record of the student’s patterns of error and see if some improvement occurs. Create accountability.  Stay on the case.

There’s a crucial distinction between revision and copy-editing. Where there are problems with sentence structure, punctuation, hyphenation, capitalization, verb forms, use of quotation marks, etc., that do not interfere with my understanding of the writing, I would tend to address those issues separately from the evolution of a paper through various drafts. In general, the closer the paper is to being done, the more attention I would pay to all of the above correctness issues (see Writing Process and Teaching Process). It makes sense to expect papers to be copy-edited to the maximum of the student’s ability before they are considered finished. Some instructors take the position that they will not record a grade until a paper appears in cleanly copy-edited form.

In general, writing instructors spend little in-class time on matters of copy-editing, partly because they usually don’t feel entirely confident about their own command of grammar. There is, however, a better reason: though many errors are made in copy-editing, few of them are made universally. Students who don’t need instruction on, let’s say, the semi-colon hate and despise sitting through a lesson on it. Therefore, these matters are usually better dealt with individually, either by written comments or in conference.

In the small minority of cases in which students write sentences that a reader truly cannot understand, sentence structure becomes the first priority. This is always a case for individual instruction, generally meaning tutoring.


For a pragmatic, clear, research-based approach to this topic, see Teaching Grammar in Context, by Constance Weaver (Boynton/Cook, 1996), especially the following:

How language is acquired: a summary                                    p. 56

Chapter 4: Toward a Perspective on Error                               p. 58

Chapter 5: Reconceptualizing the Teaching of Grammar             p. 102

Summary of the research                                                     p. 179

Glossary of Grammatical Terms                                              p. 243

In 1988, Robert Connors and Andrea Lunsford analyzed the frequency of errors in 3000 college composition papers and came up with a list of the top 20 kinds of errors. (“Frequency of formal errors in current college writing, or Ma and Pa Kettle do research.” College Composition and Communication, 39, 395-409.) This list can be found in Weaver’s book, cited above; the article is interesting in its own right for those who wish to pursue this subject.

For a further sense of the complexity of the idea of grammar, and the vexation about it among writing teachers, see Patrick Hartwell, “Grammar, Grammars, and the Teaching of Grammar,” in College English 47 (February 1985), pp. 105-127.

For a sense of the 20th-century research on the effects of teaching grammar, see Research on Written Composition, by George Hillocks, Jr. (ERIC Clearinghouse on Reading and Communications Skills, 1986).  This book surveys the research on writing and the teaching of writing from 1963 to the mid-80’s, summarizes it, analyzes it, draws inferences and conclusions from it. “One of the strongest findings of this study and [of a similar book summarizing research of the 1940’s and 50’s] is that grammar study has little or no effect on the improvement of writing.  The same is true for emphasis on mechanics and correctness in writing.  In fact, some studies indicate that when correctness is heavily emphasized in marking papers, the quality of student writing diminishes significantly.”

A recent study (2010) done at the University of Exeter (UK) offers an interesting methodological critique of the earlier research: essentially, that researchers paid insufficient attention to the teaching when they looked at the effects of “teaching grammar.” They treated the teaching as a black box; whatever kind of instruction was done, it was all lumped together into one vague category, and this made any subsequent cause-and-effect assertions highly suspect. The Exeter researchers therefore set out to “investigate the relationship between grammar teaching and writing by designing a set of teaching resources which provide pedagogical support for teaching linguistic constructions relevant to the development of writing” (my emphasis). In this way, they based their study on the kind of teaching recommended in Weaver’s book, cited above. Result: “The research provides evidence for the first time of a positive benefit of teaching grammar when the grammar is contextualised – linked meaningfully to the writing being taught.”

For more on this study, see http://education.exeter.ac.uk/projects.php?id=410


You can download this as a Word document here: What Is Grammar?