page

40

Voices from the Middle, Volume 13 Number 4, May 2006

Smith and Wilhelm | What Research Tells Us about Teaching Grammar

What Research Tells Usabout Teaching Grammar

Michael W. Smith and Jeff Wilhelm

In Polish author Walter Gombrowicz’sfamous satiric novel Ferdydurke(1938), the following exchange takes

[EDITOR’S NOTE: For seven years, Jeff has followedmy admonition that his column in VM could onlybe one page. “Teach us more, Jeff,” I’d write to him,“but do it in one page.” Somehow he always man-aged to do just that—and he never once complainedthat all the other columnists got two or three pages!Now, in a thank you to him, I’m giving Jeff and hisresearch and writing partner Michael Smith thespace they both need to help us understand a sec-ond issue, grammar. Many thanks to both Michaeland Jeff for their work on the two articles they havecontributed to this issue; they help all of us betterunderstand complex topics.]

place in a Latin grammar class:

Teacher:There is nothing more logical than a language inwhich everything that is illogical is an exception![Grammar] is a great factor in evolution!

Galkiewicz (in great agony):Evolution, schmolution. How can it develop anythingwhen it develops nothing? How can it perfect any-thing when it perfects nothing? How can it buildsomething when it builds nothing? O god, o god, ogod!

Teacher:What’s this, Master Galkiewicz? The suffix does notperfect you? You’re telling me that this suffix doesnot perfect you? That the suffix passivi futuri of thethird conjugation does not enrich you? Come, come!

Galkiewicz:The little tail ending does not enrich me! That littletail ending does not perfect me. No, not in the least!O god, O god!

Teacher:What do you mean—doesn’t enrich you! When I sayit enriches you, it most certainly does . . . Haven’t youbeen taught style, clarity of thought, precision of ex-pression [through the study of grammar] . . .?

Galkiewicz:Nothing, nothing! No art. No precision. I’m justscared of an F. That’s all I’m scared of. So I try. But Ican’t, no I can’t! (pp. 58–59)

Gombrowicz highlights what years of subse-quent grammar research have unequivocallyproven: teaching grammar—in the sense of iden-tifying or defining parts of speech, labeling, dia-gramming, or parsing sentences—is . . . well . . .to put it in Gombrowicz’s terms, senseless. First,kids don’t learn this kind of grammar, becausethere is no practical context for using this kind ofknowledge. And even if they did learn this kind ofgrammar, it wouldn’t help them be better writers,readers, thinkers, or problem solvers. There arenumerous excellent reviews of the research thatconclusively show this to be the case.

Review of the ReviewsThe past 40 years have seen two comprehensivereviews of research on the teaching of writing.Each has made an unequivocal claim about theteaching of grammar. Braddock, Lloyd-Jones, andSchoer (1963) offered this assessment:

The teaching of formal grammar has a negligible or,because it usually displaces some instruction and prac-tice in actual composition, even a harmful effect onthe improvement of writing. (pp. 37–38)

Twenty-three years later, Hillocks (1986) offeredan equally powerful critique:

School boards, administrators, and teachers who im-pose the systematic study of traditional school gram-mar on their students over lengthy periods of time in

40_43VM_May06 4/5/06, 10:30 AM40

page

41

Voices from the Middle, Volume 13 Number 4, May 2006

Smith and Wilhelm | What Research Tells Us about Teaching Grammar

the name of teaching writing do them a gross disser-vice, which should not be tolerated by anyone con-cerned with the effective teaching of good writing.(p. 248)

There have been a number of more discur-sive reviews that have focused more particularlyon research on the teaching of grammar. One ex-cellent review appeared in the “ContextualizingGrammar” issue of this journal in March of 2001.In that review, Nancy Patterson summarized sev-eral seminal studies that can help teachers to thinkabout the usefulness of various techniques of teach-ing grammar and language usage. Patterson con-cludes her review by arguing, “What all thisindicates is that the teaching of grammar musthappen within the context of larger lessons andexperiences with written and spoken language.This means that English teachers must move be-yond viewing grammar as a set of rules and a codeof correctness, and they need to re-think the ideathat a comprehensive knowledge of grammar ter-minology and rules somehow translates into aknowledge of linguistic structure or into an abil-ity to write well . . . (Patterson, 2001, 54–55).

Smith, Cheville, and Hillocks (2006) offersimilar judgments of the research on teachinggrammar. But they also point to areas of researchthat seem to offer more promise. One is the gram-matical analysis of the kinds of structures studentsneed to master in order to produce particular kindsof writing. For example, Schleppegrell (1998) dis-cusses using grammatical analyses of the structuresrequired in science writing as a way to help stu-dents develop the ability to produce those struc-tures. Another promising approach Hillocks andhis colleagues identify is sentence combining. Sen-tence combining asks students to combine kernelsentences, either in any way they choose or to pro-duce a particular structure. Strong (1973), who hasdeveloped a corpus of instructional materials us-ing sentence combining, defines the kernel sen-tence as “the basic stuff of sentences,” and providesan illustration through the following three kernels:

The writer is young.The writer is developing.The writer works with options. (p. 4)

These kernels could be combined in a number ofways, for example, “The young, developing writerworks with options,” or “The writer who is youngand developing works with options.”

The research on sentence combining clearlyestablishes both that it helps students write morecomplex sentences and that it helps students writepapers judged to be better when considered holis-tically. Strong (1986) explains this finding by not-ing that “SC may help with automaticity in syntax,freeing up mental energy so that learners can con-centrate on planning and composing” (p. 3).

However, when most teachers think aboutteaching grammar, they think about it not so muchas a way to improve students’ writing but ratheras a way to help students reduce errors. Unfortu-nately, teaching grammar by identifying or defin-ing parts of speech, labeling, diagramming, orparsing sentences doesn’t help students write morecorrectly. The research base is clear and compel-ling on this point.

Teacher Research/ReflectiveTeaching InnovationsSo where can teachers turn for models of gram-mar/usage instruction that might address the causeof error and help students to inquire into the craftof language use in ways that have an immediatefunctional value in their own writing?

One of the most compelling discussions we’veread is Mina Shaughnessy’s Errors and Expectations(1977), a book written in response to the cry overthe open admission policies a number of collegesadopted in the 1970s. Shaughnessy’s book is sorich that we find it difficult to offer a summary,but a number of principles she addresses have beenespecially influential in our own teaching. Mostsignificantly, Shaughnessy establishes that errorshappen for reasons. And if that’s so, it’s impos-sible to help students overcome them simply bynoting their existence. For example, some errorshave their roots in phonology, that is, in studentstrying to reproduce in writing what they think theyare hearing. A classic example is writing “couldof” for “could’ve.” Another common source of

40_43VM_May06 4/5/06, 10:30 AM41

page

42

Voices from the Middle, Volume 13 Number 4, May 2006

Smith and Wilhelm | What Research Tells Us about Teaching Grammar

errors is that inexperienced writers often don’tbehave the way that experienced writers do. We’veseen this especially in the tendency of some of ourstudents not to reread what they’ve written as theyare writing, a tendency that results in studentsadding on fragments rather than consolidatingsentences, as in “I watched the Eagles on Sunday.Which is my favorite team.” Another commonsource of error is language interference. Speakersof Chinese will have trouble with articles and withauxiliary verbs. Speakers of African American Ver-nacular English will have trouble with the sub-ject–verb agreement of formal English. They’llhave these troubles not because they’re carelessor lazy but because their home language eitherdoesn’t require some of the structures formal En-glish requires or because their home language pro-duces those structures in different ways.

Another source of error is hypercorrectness.Anyone who has ever taught a lesson on apostro-phes knows how likely it is that some students willput apostrophes on every word that ends with ans. Yet another is lack of confidence. Students whoaren’t confident in what they are saying rely oncircumlocutions to tentatively advance their ideas.The less straightforward a writer is, the more com-plex his or her writing is likely to be. (Our favor-ite example came from one of Michael’s students

some years ago: “I think, though I am probablywrong, that. . . .” ) This is not an exhaustive list ofthe sources of error that Shaughnessy discusses,nor is her list a complete list of the sources of er-ror we’ve identified in our secondary school stu-dents, but the awareness that errors happen forreasons has been critically important for us.

As Shaughnessy (1977) explains, recognizingthe cause of an error helps teachers both get a senseof the cost of correcting the error and the approachthey’ll have to take. An error that has its root inphonology is likely to be easier to correct thanone that’s a function of language interference, forexample. Shaughnessy also offers the powerful re-minder that while errors matter, they probablydon’t matter as much as we think they do. Shereminds us that at the very least, we need to attendto what students say as well as to how they say it.

Connie Weaver has also influenced our think-ing. Weaver is the most significant voice in thefield calling for contextualized grammar instruc-tion, and her work is essential reading. In Teach-ing Grammar in Context (1996), Weaver builds on20 years of her own work in grammar instructionto provide a rationale, theoretical justification, andinstructional models for constructivist grammarteaching in the context of students’ own writing.Her subsequent edited text, Lessons to Share: OnTeaching Grammar in Context (1998), features thework of 18 teachers who offer innovative and prac-tical insights about their experiments with thecontextualized teaching of grammar.

Harry Noden’s teaching experiments are fea-tured in Image Grammar: Using Grammatical Struc-tures to Teach Writing (1999). Noden asserts thatgrammatical structures are the tools that a writeruses to “paint” images and meaningful experiencesfor readers. His argument is reminiscent of BarryLane’s (1993) contention that “Writers don’t needto be given formulas; they need to be shownpossibilities. Then, like painters receiving newcolors, they can embrace their craft” (p. 40). Eachchapter in Noden’s book features a section called“concepts” that explores how professional writershave used particular grammatical tools of “imagegrammar” to make meaning. A corresponding sec-

CRITICAL POINTS ABOUT TEACHING GRAMMAR

1. Develop a hierarchy of grammar and usage issues based on:

✓ The commonness of the problem

✓ The extent to which it interferes with understanding

✓ The extent to which it undercuts the authority of the writer

✓ The cost of correcting it in terms of time and students’ self-image

2. Address issues one at a time, spending enough time with eachto make sure that students have mastered it before moving on.

3. Focus on identifying the cause of the error and direct instruc-tion to address that cause.

4. Teach only the terms that students need to know and onlywhen they need to know them.

40_43VM_May06 4/5/06, 10:30 AM42

page

43

Voices from the Middle, Volume 13 Number 4, May 2006

Smith and Wilhelm | What Research Tells Us about Teaching Grammar

tion of each chapter, called “strategies,” explicatesNoden’s use of classroom lessons to help studentsdevelop this same facility in their use of writing-based grammatical knowledge. Noden thus pro-vides an innovative look at how to integrategrammar instruction creatively and meaningfullyinto students’ composing processes.

Another favorite of ours is Edgar Schuster’sBreaking the Rules (2003). Our own methods stu-dents have greatly enjoyed this refreshing book.Based on his own classroom research and reflec-tive teaching, Schuster makes a case that much ofwhat we teach under the guise of “grammar” ac-tually constitutes “mythrules” that exist and areenforced only in school. He raises the issue that“Our formulas—use topic sentences, never usepassives, vary sentence openings—deny studentsthe resources routinely used by good writers, andour concentration on correctness puts the empha-sis on the least important aspect of good writing”(p. 92). Schuster provides an illuminating histori-cal discussion of how the study of English gram-mar, based on Latin, came to be a staple of schoolinstruction, and he explores the negative effectsof this emphasis. With great humor and the au-thority of his own classroom practice, Schusteradmits that there certainly are bedrock rules oflanguage use that must be honored. Though hecritiques much classroom practice, he also offersa case for appropriate foci of usage instruction andappropriate ways to implement instruction thatwill “liberate” students into the engagement ofmeaning-constructive writing with clarity, correct-ness, and creativity. Of great interest are his dis-cussions about why students may have difficultylearning particular punctuation marks or othergrammatical concepts.

These texts offer a rich set of alternatives tothe traditional teaching of grammar. They wouldoffer welcome relief to students like Gombrowicz’s

Galkiewicz by providing models of instruction thatmake sense to students and make a difference totheir writing. And that, after all, must be the ulti-mate purpose behind any teaching of grammar orlanguage use.

ReferencesBraddock, R., Lloyd-Jones, R., & Schoer, L. (1963). Re-

search in written composition. Champaign, IL: NCTE.

Gombrowicz, W. (1938). Ferdydurke. New Haven: YaleNole Bene.

Hillocks, G. (1986). Research on written composition: Newdirections for teaching. Urbana, IL: NCTE/NCRE.

Lane, B. (1993). After the end. Portsmouth, NH:Heinemann.

Noden, H. (1999) Image grammar: Using grammaticalstructures to teach writing. Portsmouth, NH:Boynton/Cook.

Patterson, N. (2001) Just the facts: Research and theoryabout grammar instruction. Voices from the Middle,8(3), 50–55.

Schleppegrell, M. P. (1998). Grammar as a resource:Writing a description. Research in the Teaching ofEnglish, 32, 184–211.

Schuster, E. (2003). Breaking the rules: Liberatingwriters through innovative grammar instruction.Portsmouth, NH: Heinemann.

Shaughnessy, M. (1977). Errors and Expectations. NewYork: Oxford University Press.

Smith, M., Cheville, J., & Hillocks, G., Jr. (2006). “Iguess I’d better watch my English”: Grammar andthe teaching of English language arts. In C.MacArthur, S. Graham, & J. Fitzgerald (Eds.),Handbook on writing research (pp. 263–274). NewYork: Guilford.

Strong, W. (1973). Sentence combining: A composingbook. New York: Random House.

Strong, W. (1986). Creative approaches to sentencecombining. Urbana, IL: ERIC/NCTE.

Weaver, C. (1996). Teaching grammar in context.Portsmouth, NH: Boynton/Cook.

Weaver, C. (1998). Lessons to share: On teaching gram-mar in context. Portsmouth, NH: Boynton/Cook.

Michael W. Smith is professor of education at Temple University and a fellow of the NationalConference on Research in Language and Literacy. He can be reached at mwsmith@temple.edu.

Jeffrey D. Wilhelm is professor of English education at Boise State University and the director ofthe Boise State Writing Project. He can be reached at jwilhelm@boisestate.edu.

40_43VM_May06 4/5/06, 10:30 AM43