September 2007
Volume 11, Number 2

Contents  |   TESL-EJ Top

Towards More Context and Discourse in Grammar Instruction

Marianne Celce-Murcia
University of California, Los Angeles and
American University of Armenia


This paper first discusses why the sentence-level drills still being used extensively in the teaching of grammar to second language learners have not been successful. What follows is a presentation of an innovative approach; namely, using context and discourse to present and practice grammar in more authentic and effective ways.

Statement Of The Problem

At the 2007 TESOL Convention in Seattle, Washington, I was surprised to see how many of the ESL/EFL textbooks being displayed and sold in the book exhibits still use decontextualized sentence-level exercises for most of each grammar lesson or unit. In such grammar exercises, learners are typically asked to do something resembling one of the rather mechanical drill types below:

  1. Pick the correct item to fill a blank given two or more choices.
    For example: I ____________ to school every day. (a. drives b. drive c. driven)
  2. Fill in the blank using the proper form of the word indicated in the parentheses.
    For example: John _______ to the store yesterday. (walk)
  3. Change a statement into a yes-no question (a) or an affirmative statement into a negative statement (b).
    For example:
    a. John is a teacher. → Is John a teacher?
    b. John is a teacher. → John is not (isn't) a teacher.
  4. Put a scrambled list of words into the right order so they form a sentence:
    For example: (my, Mr., teacher, is, Johnson) → My teacher is Mr. Johnson.
  5. Place a specified word into its correct position in a sentence.
    For example: I go to see horror movies. (never) → I never go to see horror movies.
  6. Answer a question with a complete statement.
    For example: Where were you born? → I was born in Chicago.

I could continue listing such drills, but I think I've made my point: Drills like these are neither meaningful nor authentic. Some readers might feel that the exercise shown in (6) above begins to resemble communication. However, I would argue that such an exercise lacks authenticity and does not model typical communication.

First, a question like "Where were you born?" is not asked out of the blue. It is often part of an ongoing conversation with a biographical focus, where interlocutors are getting acquainted. Alternatively, such a question can be prompted by the speaker's accent, which the listener perceives as different from his and which is not readily identifiable. The question could also be part of a structured interview, where the interviewer has a list of information items to elicit from the interviewee, for example, date of birth, place of birth, current occupation, etc.

Secondly, such a question is very rarely answered with a full statement. I asked five English speakers this question and got these responses:

  1. in St Paul
  2. South Africa
  3. on Staten Island
  4. Boston
  5. Japan—my father was a missionary

No one answered with a complete sentence. The person with the longest response elaborated because she knew her response (i.e., Japan) was an unusual one, and she thus tried to provide additional clarification.

Very few English grammar rules are strictly sentence-internal decisions that English speakers need to make. The following is a fairly exhaustive list of such mechanical sentence-level rules:

  1. determiner-noun number agreement (e.g., These books are mine.)
  2. verbs becoming gerunds after prepositions (e.g., We look upon reading books as an enjoyable activity.)
  3. reflexivization of objects referring back to subjects (e.g., Sue cut herself.)
  4. subject-verb number agreement (e.g., John walks to school.)

Even though these rules are sentence-internal, learners need to be able to apply them in the course of producing spoken or written English discourse, and we know that this is not always easy for them to do quickly and accurately.

In contrast, most of the grammatical choices English users make depend on an array of contextual factors:

  1. the interlocutor(s)
  2. the situation
  3. prior discourse
  4. shared knowledge
  5. speaker intention/purpose/stance
  6. the topic
  7. modality (speech, writing, e-mail)
  8. register (formal or informal), etc.

There is no way that sentence-level drills can give learners sufficient context to learn when and why to use the passive voice, the definite article, the present perfect tense, a relative clause, or any other grammar objective not on the short four-item sentence-level list itemized above. Only learning activities that are richly situated and fully meaningful and contextualized will begin to be able to achieve this.

Furthermore, we need to recognize that individual sentences presented in isolation are typically ambiguous in terms of their situational meaning and function. For example, the sentence "I'm hungry" means not only "speaker claims to feel hunger pangs," but it will also have different interpretations depending on the context:

1. Spoken to his mother by a youngster coming home at noon, it is a request for lunch.
2. Spoken to a passerby by a beggar with an outstretched hand, it is a request for money.
3. Spoken by a child who has just finished a meal, it is a request for more food.

From Celce-Murcia & Olshtain, 2000, p. 20)

Discourse-Based Solutions

What are some of the characteristics of grammar exercises that would better serve the needs of English language learners?

  1. If some manipulative work is needed as a warm-up, at least make it meaningful, contextualized, and reasonably authentic in terms of use.
  2. For example: Practice negation in the context of correcting false statements
    A: I just found out that Juan is from Panama.
    B: He's not from Panama. He's from Colombia.
    A: Oh.
  3. If use of a grammatical form depends on prior context as it does when using pronouns to refer back to antecedents, be sure to provide enough context so that this is clear to the learner. For example:
  4. A: What's up?
    B: I'm looking for my (purse/car keys), and I can't find (it/them). Have you seen (it/them)?
    A: No, I haven't.

    Notice that such short dialogs also contextualize practice of several tenses (simple present, present progressive, present perfect and the two negative forms can't and haven't).

  5. Find authentic texts that provide salient tokens of the grammatical form that you want to present to learners (in preparation for subsequent practice).
  6. For example: Use an e-mail message to present the fact that future scenarios are often initiated with the "be going to" future and subsequently elaborated with "will/'ll". After discussion and analysis of the future forms, the learners can write their own future scenarios and e-mail it to classmates with a copy to the instructor. Here is a sample for analysis and discussion:

    Hi Sue!

    How are you? I hope you're fine. Guess what? I'm going to sing in the mixed chorus this year. I'll have practice sessions on Wednesday evenings, and we'll prepare pieces for several concerts and events during the year. We'll even travel to Washington for a choral competition. It'll be fun. What's new with you?

    -Best, Sally

    (See Celce-Murcia & Larsen-Freeman, 1999, chapter 9 for further data)

  7. Grammar instruction can be integrated with tasks designed to prepare learners to read and write academic discourse. For example, since the formal report genre lends itself to the use of passive voice, one can facilitate practice of the passive in the context of pair work, where partners are told to survey five people each, combine their results, and write a short report based on an example text.
  8. Survey question: What is your favorite movie/book/city/, etc.

    Example report: Ten Americans were asked, "What is your favorite city?"

    Three cities in the U.S.—New York, San Francisco, and Boston—were mentioned twice, while Seattle was cited once. London, Paris, and Rome were each selected by one survey participant.

On a related note, for teachers who say that they have tried to do grammar correction using sentences from students' essays, but that it didn't seem to help much, the following strategy is proposed in lieu of doing sentence-level correction. For example, to practice correcting verb forms that have been covered in class, present an exercise such as the following so that learners get accustomed to correcting errors in coherent text (not just in sentences), which in turn better prepares them to edit their own work.

Directions: Work with a partner to correct all the verb form errors in this passage. There are six.

Passage: Personal Digital Assistants are become very popular now. They not have a keyboard or a mouse. Most PDAs doesn't have word processors, spread sheets, or databases. But PDAs has a datebook, a clock, a calculator, and a notebook. You can even go on the Internet with some PDAs. People sends and receives e-mails with these PDAs.

(text from O'Sullivan, 2007: 219)
(With some learners, it is not necessary to give the number and type of error. The teacher should adjust the level of difficulty to the class.)


I conclude by emphasizing that grammar instruction is much more effective when it is situated in a meaningful context, embedded in authentic (or semi-authentic) discourse, and motivated by getting learners to achieve a goal or complete an interesting task. Hopefully, at future TESOL conferences, we will see more materials for grammar instruction that satisfy these criteria with a concomitant decrease in the quantity of materials consisting primarily of manipulative sentence-level grammar drills.

About the Author

Marianne Celce-Murcia is Professor Emerita of Applied Linguistics at UCLA and Dean of English Programs at the American University of Armenia, Yerevan. Her long-standing interest in English grammar and grammar pedagogy have resulted in (among other publications) her co-authoring The Grammar Book (2nd ed. 1999) with Diane Larsen-Freeman and co-editing The Grammar Connection (2007, a grammar textbook series) with Maggie Sokolik (both publications with Thomson Heinle).


Celce-Murcia, M. & Larsen-Freeman, D. (1999). The grammar book: An ESL/EFL teachers course. 2nd edition. Boston: Thomson Heinle.

Celce-Murcia, M. & Olshtain, E. (2000). Discourse and context in language teaching. New York: Cambridge University Press.

O'Sullivan, J.K. (2007). Grammar connection: Structure through context, Book 1 (series eds., M. Celce-Murcia & M.E. Sokolik). Boston: Thomson Heinle.

Comment/view comments on this article. There are currently comments.

© Copyright rests with authors. Please cite TESL-EJ appropriately.

Editor's Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations.