The Language Teacher

Grammar FROM Context: Re-thinking The Teaching Of Grammar At Various Proficiency Levels Patricia Byrd Georgia State University, Atlanta, Georgia USA patbyrd@gsu.edu

Return to The Language Teacher Online

Current Problems With Systems for Organizing Grammar in the ESL/EFL Curriculum

Around the world, EFL and ESL students and teachers are being frustrated because of the continued use of an approach to the integration of grammar into programs, classes, and materials that has failed us badly. This traditional approach divides up grammar in a system that ignores the nature of English and of authentic communication using English. Table 1 gives the grammar curriculum still in place for a sequence of composition courses taught at a large community college ESL program in the U.S. The system, however, is not unique to that institution; similar grammar divisions are found around the world in many different types of educational settings and are the backbone of many textbooks and curricular plans. Other systems scatter verb tenses over the various proficiency levels. For example, Level 1 works with simple present and present progressive and a couple of the modal auxiliaries (will and can). Level 2 gets to learn about past tense and a few more modals. Level 3 studies present perfect and the rest of the modals and even some semi-modals (be able to and ought to). Passive voice is reserved for Level 4.

Table 1

A Traditional Divide-Up-the-Grammar Curriculum Statement for an ESL Composition Sequence

Level 1

• simple sentences (affirmative & negative)
• compound sentences (and, but, or, so)
• complex sentences (while, before, when, because, after) [optional for this level]
• simple questions (yes, no)
• information questions (who, what, when, where)
• nouns (singular & plural)
• pronouns (objective, demonstrative)
• possessives (noun, pronoun, adjective)
• position of adjectives (comparatives and superlatives) [optional for this level]

Level 2

• 75% mastery of level 1
• simple sentences
• compound sentences
• complex sentences with time clauses
• simple questions
• information questions (which, whose, whom, when, because)
• noun phrases (count, noncount with articles)
• pronouns (reflexive and impersonal "you")
• adjective clauses (restrictive) [optional at this level]
• noun modifiers

Level 3

• 75% mastery of level 2
• simples sentences
• compound sentences
• complex sentences with time and cause-and-effect clauses
• simple questions
• information questions
• tag questions
• nouns (collective and abstract)
• adjective clauses (restrictive)
• adjective clauses (non-restrictive) [optional at this level]

Level 4

• 75% mastery of level 3
• simple sentences
• compound sentences
• complex sentences
• simple questions
• information questions
• embedded questions
• noun clauses (reported speech)
• adjective clauses (reduction)--participles as modifiers

Easy Grammar vs. Hard Grammar: Inconsistent Decisions

Such systems for dividing the grammar among proficiency levels have often been based on ideas about the easiness or difficulty of certain structures for learners. The lack of consistency in the decisions that have been made using these standards suggests that the "ease" and "difficulty" cannot be defined with certainty. Indeed, some lack of clarity about the differences between "easy/hard to learn" and "easy/hard to teach" have added to problems with such systems (see, < http://www.gsu.edu/~eslhpb/grammar/difmatrx.htm> for a discussion of an ease/difficulty matrix). Crandall (1985) discusses another mistaken notion about grammatical difficulty in the commonly held idea that the passive is an "advanced" form that must wait until quite late in an ESL curriculum; she found that students even at the very earliest stages of elementary school learn such phrases as "4 is divided by 2." That is, whatever is going on with passive voice, children who are learning ESL to participate in U.S. elementary education cannot wait around to learn the passive voice when they are at a "more advanced level."

Problems Caused for Teachers by the Grammar Curriculum

A common complaint voiced by teachers working in traditional systems such as that illustrated in Table 1 is "I can't cover all the material. If I don't, the students won't have another chance. But my students just can't learn all of this material in one term!" A peculiarity of the traditional design is its assumption that students can "put off" elements of grammar for later in their lives--notice how the curriculum in Table 1 delays the use of noun phrases and articles for Level 2 and demands mastery (or 75% mastery) at that Level. The frustration of teachers with such a plan is only natural--such a division is both inauthentic (students can't wait until Level 2 to start encounters with noun phrases since any authentic English will have articles and noun phrases) and unrealistic (few students will "master" noun phrases in a single term of study). Very often, teachers find that they need to teach grammar that is supposed to be delayed to a higher level; they also find that students do not yet have "mastery" of grammar that they were supposed to have learned at a lower proficiency level. As we know from second language acquisition research (for example, Ellis, 1994, especially chapter 9), students cannot be expected to learn everything correctly through just one presentation during just one academic term; most learners need repeated "spiraling" of a topic and of various aspects of a topic. While most teachers realize the importance of such patterned repetition for ESL/EFL students, spiraling of grammar cannot easily be achieved if the curriculum does not allow for it.

Finding Solutions by Looking at the Nature of Grammar in Use

The Challenge of Using Authentic Materials and Tasks

Around the ESL/EFL world, we hear calls for the use of "authentic" tasks, materials, and activities in our programs and courses. While the term has many different interpretations, its wide appeal seems to be based on the recognition by teachers that students can learn to pass courses based on made-up English without learning to communicate in English outside of the classroom. At the same time, the use of authentic materials is a substantial challenge for most curricula and for many teachers because authentic materials often include vocabulary and grammar that are supposed to be "too difficult" for students "at this level." To select and use authentic materials, we have to break with long held beliefs about language proficiency, about language proficiency levels, and about the nature of English itself.

Misconceptions about Language Behind Grammar Curriculum Designs

Divide-up-the-grammar approaches to curriculum and materials design misconceive the nature of language in two ways: First, these schemes build on de-contexualized abstractions about language. They assume that all features of English are used in the same ways in all settings. We know both from research and from personal experience of using English that this is not accurate: An informal chat with a friend about where to go to dinner will use different grammar and vocabulary from a research report on the nutritional value of fast food.

Secondly, and as importantly, research on the nature of language in use for authentic communication in real world settings demonstrates that the older system separates features of language that actually belong together when they are used in context. Researchers studying genre and discourse have shown that different types of communication have different grammatical signatures; that is, different types of communication make use of different grammatical structures and can be characterized in terms of those grammar items (Biber, 1988; Biber, Conrad, & Reppen, 1994; Conrad, 1996a & 1996b; Grabe, 1987). It turns out that grammar has a lot in common with vocabulary as revealed in computerized studies of extensive collections of written and spoken English. We have learned about collocation as a feature of the words in a language: certain words tend to be used together in authentic communication (Sinclair, 1987a, 1987b, 1991). Indeed, one of the features of a "foreign accent" is using combinations of words that have the correct meanings but that native speakers just would not use together. Similarly, grammar structures tend to cluster together in particular chunks in particular types of communication.

Looking at the Grammar of Authentic Samples of Written English

Consider the grammar that we can see in the two small samples of English in Samples 1 and 2. These are both authentic writing that was published in the U.S. and intended for general rather than specialist readers. They are used here to show the grammar of "real" written English as it is written by writers who are not thinking about teaching ESL/EFL but about communicating content to readers.

Sample 1 Narrative in a Historical Study of the State of Alabama for Experienced Readers De Soto Against Tascaluza from Chapter Two "Exploration and Colonization" Alabama: The History of a Deep South State On the morning of October 18, 1540, the entourage entered Mabila, a village located on a plain above a wide river. The town was surrounded by a palisade, and inside were eighty large houses fronting a square. Tascaluza disappeared inside a house, and the Indians began to dance and sing while the Spanish grew more suspicious and uneasy. Suddenly, the Indians attacked, shooting arrows from the houses and forcing the Spanish to flee the village, leaving some of their horses behind. The Indians promptly killed the feared animals. De Soto rallied his men for a counterattack and set fire to the village. The battle, most of it hand-to hand combat, lasted until nightfall. In his manuscript, the Inca claimed 11,000 Indians died; Biedma reported 5,000 killed; and the more reliable Ranjel related that the Spanish found 3,000 Indian bodies without counting the dead inside the burned village. The Gentleman of Elvas reported 2,500 dead. Sample 2 Narrative in the History of Science for Young Readers from "Gregor Joann Mendel: The Mystery of Heredity" Break-Throughs in Science In 1900 three strangers met at a crossroads of research. Each, without knowledge of the other two, had worked out the rules that govern inheritance of physical characteristics by living things. The three were Hugo de Vries of Holland, Carl Correns of Germany, and Erich Tschermak of Austria-Hungary. Each made ready to announce his discovery to the world. In preparation, each looked through previous issues of various scientific journals, to check earlier work in the field. Each, to his astonishment, found an amazing paper by someone named Gregor Johann Mendel in a 35-year-old copy of an obscure publication. Mendel, in 1865, had observed all the phenomena that the three scientists were preparing to report in 1900. Each made the same decision. With the honesty that is one of the glories of scientific history, each abandoned his own claims and called attention to Mendel's discovery. Each man advanced his own work only as confirmation.

Sample 1 was written for an adult audience with advanced literacy skills. Sample 2 was written for young readers with less advanced literacy skills. While both of these narratives "tell a story," neither is written simply to tell the story; both are given to illustrate a principle. These two samples of narrative, each intended for quite different audiences at different "proficiency levels," have much in common grammatically: (1) past tense verbs, (2) chronological organization often signaled by time adverbials, (3) proper nouns, and (4) use of 3rd person personal pronouns (to refer to the people named with the proper nouns).

What Makes for Proficiency Level Differences?

Here is the challenge of such an analysis: These two samples use the same basic grammatical structures, yet they were written for very different audiences and seem to represent different levels of reading proficiency. If they are grammatically alike, then what makes for the differences? Certainly the difference is not in the length of the samples--both are in the 150-word range. Schema theory might help explain the difference for many readers--some familiarity with the Spanish exploration of the new world will help to understand the first sample; some familiarity with the history of science/biology will help with the second. Those differences in content are valuable in making proficiency level decisions because prior knowledge can make a reading passage more accessible for learners while current interest in a topic can lead to the motivation that carries learners along even when a passage is difficult for them.

The samples do hint at possible linguistic differences that involve discourse level rather that sentence structure complexities: the first sample uses more characters in a more complicated story while the second simplifies the story line down to the actions of "each scientist." The second sample is capable of standing alone without a great deal further contextualization; the first tells a complete "story" but leaves the reader in need of additional information that is not provided here (who are these people making counts of dead Indians? Why were the horses "feared"?) Even the sentence types suggest discourse level decisions about style (and perhaps audience) rather than simply reflecting sentence-level choices: The first sample uses a number of participle phrases ("located on a plain above a wide river," "fronting on a square," "shooting arrows from the houses," "forcing the Spanish to flee the village," "leaving some of their horses behind") in contrast to the somewhat simpler sentence structures of the second sample. The first sample also has a high proportion of compound-complex sentences--a decision by the authors that might be based on disciplinary preferences for this style (see MacDonald, 1992, for more information on writing in the humanities and social sciences in the U.S. academic setting).

Grammar Curriculum Design Based on Authentic Materials and Real World Tasks

Rather than starting our grammar design with a list of grammatical structures, we need, when possible, to begin with an analysis of the uses to which our students will put their English--the communicative purposes for their study of ESL/EFL (Byrd & Reid, 1998; Carson, Chase, Gibson, & Hargrove, 1992; Long and Crookes, 1992). Knowledge about their real world communication tasks gives us access to realistic, authentic samples of the language used in such tasks. This combination of task-based and genre-based curriculum design has been used extensively in Australia (Hyon, 1996) and is gaining popularity in the U.S., especially in ESL programs that have clearly defined purposes such as English-for-academic purposes or English for business communication. Once we are clear about the purposes of our students, we can turn to the growing literature on discourse/text/genre analysis to seek more information about the language that our students need to learn to use effectively (for example, Bardovi-Harlig, 1996; Meyer, 1996).

Difficulties of Identifying Student Goals in Many Settings

However, many students are taking English courses for less-than-clear reasons--because of requirements in their academic programs, because their friends are taking it, because their parents think it a good idea, because it might help get a better job, and other vague if worthy purposes. For these students, we as teachers must make "educated guesses" based on our experiences of the ultimate uses to which other similar students have put their English. In addition, we now have research that reveals the grammar system that lies behind genres and tasks. The analyses of written and spoken English provided by Biber (1988), Grabe (1987), and other of their colleagues (Biber, Conrad, and Reppen, 1994, and Conrad 1996a & 1996b) show the systematic intertwining of sets of grammar structures in various settings. Biber (1988) gives a detailed explanation of the research methodology and of his results; Conrad (1996b) provides an accessible explanation of the multidimensional system.

Value of Discourse Approach for ESL/EFL

The fundamental value of this system for ESL/EFL teachers, curriculum designers, and materials writers is that it gives us a map to use in making decisions about what grammar to teach and in what combinations. If we know the ultimate purposes of our students, we can select texts and tasks based on those purposes--and can present the grammar in a rational, coherent pattern that is not limited to its use just in a particular type of communication. If we and our students are unclear about their ultimate purposes, then we can be sure that they have experience with an appropriate range of English communication and linguistic patterns as a background to future uses that we cannot now anticipate.

In Variation across speech and writing (1988), Douglas Biber gives detailed results from his computerized analysis of a large collection of spoken and written English. For our purposes here, a major value of this study is the information that it provides about the grammatical features that combine to make up different genres. Biber identifies six different linguistic dimensions that can be combined in various proportions to characterize different samples of speaking or writing. For example, the sample of U.S. history in Sample 1 has many of the features of past time narratives--past tense verbs, chronological organization, and 3rd person personal pronouns to refer to characters identified with proper nouns, but this type of discourse also has features of more abstract academic writing with its long complicated noun phrases, the passive sentence, and the turn from chronological to logical order in the final part of the passage.

Differences and Similarities between Conversation and Writing

The first two of Biber's dimensions are especially powerful tools for the analysis of language-in-use and for moving toward a more authentic, realistic, and practical system for organizing the grammar in ESL/EFL curricula or materials. Dimension one involves the contrast between conversational communication (called "involved" by Biber) and more technical, academic communication (called "informational" by Biber). The second dimension illustrates the importance of past time narrative and the contrast between the language of past time narration and other communication in English. A combination of these two dimensions provides a picture of eight different ways in which grammar is clustered for particular types of communication. For example, informational narratives are often found in academic writing about history (as in Sample 1). Another example combines some of the features of conversation with informational writing--as when a textbook writer structures explanations in a series of questions and answers, uses contractions, and addresses the reader as "you."

Putting Modal Auxiliaries and Conditional Grammar in Context

Another important observation in the work by Biber, Grabe, and Conrad is the use of modal auxiliaries and conditional sentences (with if-clauses) in writing that is trying to be "persuasive." For ESL/EFL teachers, this stands our traditional approach on its head--or at least turns our heads in a better direction. Rather than teaching students about conditional sentences as if the main issue involves getting the right verb tenses in the right clauses, we need to teach students that they are trying to build combinations that will convince a reader of the truth of their argument. The persuasive purpose comes first--then getting the tenses right is part of the larger obligation of the writer to take meaning and audience into consideration. The same re-focusing is true for uses of modal auxiliary verbs to persuade a reader that the writer is being reasonable and judicious. For example, rather than the bald generalization

"Eating hamburgers and french fries for lunch everyday causes high blood pressure and obesity."

A more acceptable generalization might possibly be

"Eating hamburgers and french fries for lunch everyday could cause high blood pressure and obesity."

Applications of Discourse Research in a Re-Design of the Grammar Curriculum

While a complete program would need to consider all the dimensions proposed by Biber (and his colleagues in publications such as Conrad, 1996a and 1996b), we can begin our re-thinking of the curriculum by focusing on the grammatical characteristics that underlie English communication.

Three Principles to Guide Re-Design of the Grammar Curriculum

Much research and analysis remains to be done. We need to know details about the nature of authentic materials and activities that are appropriate for learners at different stages of language acquisition. We need to know more about the details of the "grammar clusters"--those grammatical structures that generally occur together in a particular type of communication. For example, narratives often cluster together (1) past tense verbs, (2) proper nouns, (3) first and third person pronouns, and (4) adverbs of time and place.

We need systems for testing and assessment that are built on grammar in clusters rather than individual bits of structure. While we do not yet know all the details of such clusters, the information provided by discourse studies gives us a basis on which to re-think our approach to grammar in ESL/EFL programs and materials. Here are three principles that such analyses suggest to me:

1. In authentic communication, the structures of English cluster together in predictable patterns. These grammatical clusters do not change at different "proficiency levels."
   
   Therefore, the concept of "proficiency level" must be reconceived on the basis of something other than grammar structures. Even levels of "grammar" courses and "grammar" textbooks should not be based on exclusive divisions of grammar among the levels.
   
   
2. These grammar clusters involve combinations of structures that have often been taught separately: in authentic use, verbs combine with nouns and other structures to make up the grammar package used for a particular type of communication. Therefore, rather than thinking about the grammar curriculum in terms of separate features separately taught, we have to learn to think about features in combination with each other. For example, in the past we thought about the verbs as a separate system and organized presentation of verbs without thinking much about how verbs connect with nouns or other grammatical features.
   
   
3. All students at all proficiency levels need to work with the range of grammar clusters that lie behind English genres. All students need work with authentic oral communication with its use of informal vocabulary, fragments, you/I, questions, and turn-taking about everyday topics. All students need work with the features of past time narrative and its use in a wide variety of settings (not just literature but business case studies, newspaper reports, research reports, and other uses of "stories" to explain and support generalizations). All students at all proficiency levels need to learn to communicate about general topics--theories, ideas, data, and so forth--using the grammatical features that cluster together for such communication (complex noun phrases, passive voice, technical or specialized vocabulary, and so forth). All students at all levels need to learn about ways to be persuasive and convincing for different audiences. Additionally, students need to learn to see how these subcategories interact with each other--the use of conversational grammar in written settings (when writers use contractions, ask questions, refer to the reader as "you," and other conversational strategies).

Problems with a "Grammar Strand" Approach

The traditional grammar curriculum has a verb strand, a noun strand, a preposition strand, an adverb strand, and so on. For example, we thought about how present tense contrasts with past tense, but we didn't think about how past tense tends to be used with the grammatical structures such as those that we found together in the narrative samples given earlier in this discussion in Samples 1 and 2.

In systems like that shown in Table 1, the curriculum divides grammar up not just among proficiency levels but between different courses--the composition courses take nouns, clauses, and sentences while the grammar courses focus on verbs (and a few other miscellaneous structures). No wonder teachers are frustrated and students are confused. Curricular design like that suggested in the following discussion is needed to provide a practical, coherent system for the basis of whole programs, individual courses, and the materials to be used in them.

Tentative Redesign Suggestions

Tables 2 and 3 are incomplete and tentative but are provided as examples of the way that I've been thinking about grammar in the ESL/EFL curriculum as a result of my study of discourse analysis and my observation of the problems that students and teachers are having with the current approach to grammar. In this curricular model, rather than planning a verb strand or noun strand, strands are based on the communication that students really need to learn to handle and on the grammar that underlies those communicative needs. Content is based on the students' needs and interests. Genre types (letters, reports, newspapers, textbooks, quiz and test questions and answers, conversations with friends, oral reports in the business setting, and so forth) are selected from those that the students will be required to understand and produce. The grammar items are selected and organized based on the work by Biber and other discourse analysts (for example, Bardovi-Harlig, 1996; MacDonald, 1992; Pica, 1987) that shows how different grammar items cluster together in the selected genres and how different grammatical structures are used for different discourse purposes.

In this design, proficiency level differences are not based on items of grammar covered at the different levels. As we have seen, the patterns of grammar structures remains essentially the same at all levels. The differences in the proficiency levels are worked out in terms of lexical complexity and the cognitive demands of the materials being used and the activities being done by the students. Additionally, student consistency in accuracy develops over the levels--that is, one of the characteristics of "higher proficiency" is better accuracy as well as more fluency.

In this design, students at all proficiency levels work with narrative, informational texts, and conversational interactions--and with learning about being persuasive for different audiences. For example with the narrative strand, even beginning students can read and write narratives; at the next higher level, these narratives can be longer and more complex in their internal structure--with more characters to refer to and more information to share; at the most advanced level students can work with narratives used for multiple purposes.

Overcoming the Obstacles to Needed Change

Clearly, the implementation process for such a major change in ESL/EFL curricular design will be difficult. We face at least two significant problems. First, the research on grammar, discourse, genres and tasks, while substantial and mature, is far from complete. Application of the research into the grammatical structures that characterize particular written genres has been limited, but the potential of the work is undoubtedly beginning to have a considerable impact on the teaching of ESL/EFL.

The second problem that we face in rethinking and redesigning the grammar curriculum is even more important and thorny. The changes I am proposing mean a substantial curricular change in our field. Change, I know, is not fun and not easy.

The more complex the change, the more time and resources are needed to persuade others that the change is beneficial, and the more patience and work is necessary for those who will implement the change. I am convinced that we are now at one of those time of change when the problems caused by the old system are so serious and so evident that even the most reluctant ESL/EFL professional--seeing ahead all of the changes in materials, handouts, tests, and classroom activities required by a new system--will be ready join in the messy but satisfying work involved in creating an ESL/EFL grammar curriculum that that is in congruence with our knowledge about English and about language learning.

Table 2
ESL/EFL Curriculum Designed Around Grammar Clusters

Core Communication Types	Other grammar and study suggested by these features
Spoken Conversation (although much of the following would also be true for written conversation such as e-mail) 1. present tense, including be as main verb 2. contractions 3. fragments rather than complete sentences 4. questions 5. you & I (we) 6. vocabulary common to speaking such as hedges such as about, something, almost, like) and words like as a lot, for sure, really) and noises such as "ummm" and "ahhhhh."	1. speech vs. writing (turn taking, involvement with listeners, etc.) 2. answering questions (as well as asking them) 3. verb forms to understand contractions--including do/does/did with negative contractions and questions 4. subject-verb agreement (present tense and forms of be)
Communication about theories, data, information, especially reading and writing about abstract topics 1. complex noun phrases 2. longer and more complex vocabulary and more variety in vocabulary than in involved 3. prepositions 4. agentless passive sentences 5. present tense (because of heavy use in informational prose)	1. noun types and forms (regular and irregular) 2. preposition uses (especially used as post noun modifier) 3. formation of passive sentences and verbs 4. regular and irregular verbs (for the past participle of passive verb) 5. complete sentences vs. fragments
Past Time Narrative 1. past tense verbs 2. he/she/it (and their various forms) 3. aspect verbs 4. verbs such as say, report 5. present participle clauses	1. regular and irregular verbs 2. proper nouns (for use with the personal pronouns) 3. noun clauses (with say, report) 4. sentence combination (to get the participle clauses) 5. chronological organization (and time adverbs to signal it) 6. complete sentences
Features common to all written work	1. Complete sentences (vs. the fragments of speech) 2. Sentence types and variety

Table 3
Grammar at Four Proficiency Levels

Discourse Types all the grammar items in Table 2 will be worked with at the 4 levels with different vocabulary and content and different expectations for accuracy and consistency	Level 1 forms in context, observing more than required to produce	Level 2 forms in context, more production but still observing authentic uses	Level 3 more and more accurate and consistent in accuracy	Level 4 more consistency and details, expanded vocabulary, expanded ability to do larger-scale activities
Interactive Grammar for tasks such as having a conversation with peers or business colleagues, doing group work required by courses, interviewing to collect information for projects, and other tasks that require interactive communication	basic conversational skills, focus on contexualized forms, starting to understand some of the social-cultural features of conversation in English in different settings	more practice with conversation skills and opportunities to listen to and analyze "conversations" from movies, TV, and other sources--same grammar items but more sophisticated content and more complex activities	doing surveys to gather data--same grammar items but more complex content and activities and more expectation for accuracy	Using interviews to get information for research projects--fluency and accuracy in using the grammar items that make up the cluster
Past Time Narrative for tasks such as reading and writing for history courses, reading and writing in based on literature, reporting research procedures in the social sciences, responding to case studies in business, reading newspapers	telling stories about themselves and their families learning verb forms for past tense, regular & irregular verbs, other forms required by tasks	reading and writing narratives to use for more than simple story telling	study of world or area history--reading content more advanced than production but writing in response and using content in own writing	using narratives to provide examples for statements of theory--in some social science of interest or use to students--reading, analyzing, and writing case studies
Informational Grammar for tasks such as reading and writing about theory or data in the sciences or writing definitions on tests or other communication about generalizations	reading and writing about general topics, learning about definitions--observing passive if not able to produce it	learning to produce the full range of structures	able to produce generally accurate descriptions and statements of theory	reading and writing about theoretical issues appropriate for content interests
Sentences & Other Grammar Common to Other Clusters for tasks that require writers to use complete sentences or that require readers to understand the implications of different sentence types	trying out what they can do, observing more than they can produce--but generally able to write a variety of types using content and vocabulary appropriate for beginners	able to produce all forms with appropriate vocabulary and content--considerable variation in accuracy	more consistency in production and more complexity in content	Accurate formation along with a developing sense of style (using the different types for dramatic effect, etc.)

References

Bardovi-Harlig, K. (1996). Tense and Aspect in (Con)Text. In T. Miller (ed.), Grammar and Discourse. The Journal of TESOL France, 3, 19-33.

Biber, D. (1988). Variation across speech and writing. Cambridge: Cambridge University Press.

Biber, D., Conrad, S. and Reppen, R. (1994). Corpus-based approaches to issues in applied linguistics. Applied Linguistics, 15(2), 169-189.

Byrd, P. and Reid, J. (1998). Grammar in the Composition Classroom. Boston: Heinle & Heinle.

Carson, J. , Chase, N., Gibson, S, & Hargrove, M. (1992). Literacy demands of the undergraduate curriculum. Reading Research and Instruction, 31(4), 25-50

Conrad, S. (1996a). Academic discourse in two disciplines: Professional writing and student development in biology and history. Dissertation. Northern Arizona University. UMI order no: AAI9625751

Conrad, S. (1996b). Investigating academic texts with corpus-based techniques: An example from biology. Linguistics and Education 8, 299-326.

Crandall, J. (1985). The language of mathematics: The English barrier. (Project abstract for the Fund for the Improvement of Postsecondary Education ESL/Math Project). Washington, DC: Center for Applied Linguistics.

Ellis, R. (1994). The study of second language acquisition. Oxford: Oxford University Press.

Grabe, W. 1987. Contrastive rhetoric and text-type research. In Ulla Connor and Robert Kaplan (Eds.) Writing across languages: Analysis of L2 Text, pp. 115-137. Reading, MA: Addison-Wesley Publishing Company.

Hyon, S. (1996). Genre in three traditions: implications for ESL. TESOL Quarterly, 30(4), 693-722.

Long, M. and Crookes, G. (1992). Three approaches to task-based syllabus design. TESOL Quarterly 26(1), 27-56.

McCarthy, M. (1991). Discourse analysis for language teachers. Cambridge: Cambridge University Press.

MacDonald, S. (1992). Professional academic writing in the humanities and social sciences. Carbondale, IL: Southern Illinois University Press.

Meyer, L. (1996). The contribution of genre theory to theme base-based EAP: navigating foreign fiords. TESL Canada, 13(2), 33-45.

Pica, T. (1983). The article in American English: What the textbooks don't tell us. In N. Wolfson and E. Judd, Eds., Sociolinguistics and language acquisition 222-233). Rowley, MA: Newbury House.

Sinclair, J. (1987a). Introduction. In J. Sinclair (Ed.), Collins Cobuild English Language Dictionary. London: Collins.

Sinclair, J. (Ed.). (1987b). Looking up. London: Collins.

Sinclair, J. (1991). Corpus, concordance, collocation. Oxford: Oxford University Press.

Swales, J. (1990a). Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press.

Swales, J. (1990b). Nonnative speaker graduate engineering students and their introductions: Global coherence and local management. In Ulla Connor and Ann J. Johns (Eds.) Coherence in writing: Research and pedagogical perspectives, pp. 189-207. Alexandria, VA: TESOL.