The Language Teacher
December 2000

Importance of "MultiWord Chunks" in Facilitating Communicative Competence and its Pedagogic Implications

Hazel Ketko

Hiroshima International School, Hijiyama University


This article will discuss the importance of "multiword chunks" in facilitating communicative competence, by contrasting the frequency of chunk use by native speakers of English (NSs) and Japanese learners of English (JLEs) in their spoken discourse. Based on the findings, some of the pedagogic implications for "vocabulary teaching" will be discussed. Although multiword chunks occur in both written and spoken discourse, this article is concerned only with chunks in spoken discourse.

Definition of "Multiword chunks"

In this article, the term "multiword chunks" is used broadly to refer to vocabulary items consisting of a sequence of two or more words which "semantically and/or syntactically form a meaningful and inseparable unit" (Moon, 1997, p.43). They include collocations (e.g. "alcoholic drink"), polywords (e.g. "by the way"), idioms (e.g. "take action"), phrasal verbs (e.g. "put off"), fixed phrases with pragmatic functions (e.g. "no kidding"), and "sentence frames" which allow some degree of inflection and expansion (e.g. "The + (intensifier) + adjective + thing is . . ."). The terms "multiword chunks" and "chunks" will be used interchangeably in this article. In-depth discussions about the terminology and identification of multiword chunks can be found in Yorio (1980), Nattinger & DeCarrico (1992), Moon (1997), and Carter (1998).

Importance of Multiword

Chunks in Facilitating Communicative Competence

Widdowson (1989) views communicative competence as being composed of two elements: "grammatical competence" (knowledge) and "pragmatic competence" (ability). He explains that "knowledge can be characterized in terms of degrees of analyzability, ability can be characterized in terms of degrees of accessibility" (p.132). His concept of analyzability leads to the question: "How far can the English language be analyzed?" Recent research (e.g. Bolinger, 1975; Pawley and Syder, 1983; Peters, 1983) suggests that a large part of the English language is formulaic, and is based on units beyond the level of the orthographic word. In other words, a majority of the linguistic knowledge of English NSs "takes the form less of analyzed grammatical rules than adaptable lexical chunks" (Widdowson, 1989, p.132). Bolinger (1975, 1976) suggests that a large part of the English language, although analyzable, is not being used on an analyzed basis. Pawley & Syder (1983, p.193) also indicate that "native speakers do not exercise the creative potential of syntactic rules to anything like their full extent." Thus, grammatical competence is not a matter of being able to analyze a single sentence down to its every word, but rather, it is "a matter of knowing how the scale is to be applied: when analysis is called for and when it is not" (Widdowson, 1898, p.133). Over-analysis of grammatical rules will lead to production of such expressions as "Before you leap, look," which is grammatically correct but "linguistically ill-formed" (ibid., p.133)-- a sign of incompetence.

Widdowson's concept of accessibility is the ability to access the mental lexicon to retrieve a piece of "ready-made" language appropriate to a particular context. This "schematic view of competence" (Aston, 1995, p.262) rests on the underlying belief that the human memory system stores lexical representations in multiple forms (Bolinger, 1976; Pawley & Syder, 1983). This concept of accessibility of multiple representations implies an element of "choice" that has to be made during language use. This choice is what Pawley and Syder (1983) refer to as "nativelike selection," which suggests that "the language system enables particular meanings to be conveyed in multiple ways, and that these different paths do not all have equal value: some expressions seem to acquire some sort of valence" (Skehan, 1995, p.97). According to Pawley & Syder (1983, p.193), only a "small proportion of the total set of grammatical sentences are nativelike in form", and if a language user selects expressions which are not nativelike, his/her expressions will be judged to be "unidiomatic, odd or foreignisms" (Pawley & Syder, 1983, p.193).

The above brief discussion highlights the fact that knowledge of multiword chunks and how to select and use them in appropriate contexts is a sign of communicative competence. Widdowson (1989, p.135) puts it succinctly:

Communicative competence is a matter of knowing a stock of partially pre-assembled patterns, formulaic frameworks, and a kit of rules and being able to apply the rules to make whatever adjustments are necessary according to contextual demands.

A Comparison in the Use of Multiword Chunks between NSs and JLEs

In a recent study (Ketko, 2000), two research corpora (Corpora A and B) compiled for a different study (Cox, 1998) were analyzed by the author for NS chunk use in their spontaneous spoken discourse while engaging in two communicative tasks (see Appendix A). Corpus A consisted of 8,097 words and was made up of transcripts of 24 pairs of NSs doing Task A. Corpus B consisted of 8,088 words and was made up of transcripts of 25 pairs of NSs doing Task B. The NS chunk use was then compared with that of six Japanese learners of English (JLEs) while doing the same tasks. The JLEs had over 10 years of formal English language education (six years in junior and senior high school, four years in university), and were considered to have a high proficiency in the English language (they were all English majors in university and had a TOEIC score of 600 or above). However, none of them had had exposure to English outside the classroom.

The study found that NSs used significantly more chunks than JLEs: The NSs used an average of one chunk in every three to five words, whereas the JLEs used an average of one chunk in every 11 to 16 words. The study also revealed that the most common types of chunks used by NSs while doing the communicative tasks were:

  1. Fixed phrases with pragmatic functions, such as realizing communicative strategies (e.g. paraphrasing, stalling, giving evaluative response), mitigating own opinion by using "vague language" (Channell, 1994), etc. For example, I mean; you know; that's interesting; . . . or something like that.
  2. Fixed phrases with discoursal functions, e.g. of course; because of.
  3. Common collocations, including prepositional phrases and phrasal verbs, e.g. deal with; at present.
  4. Sentence frames labeled "lexicalized sentence stems" by Pawley & Syder, (1983). For example, N + would be [adj.] to + V: I would be happy to help; She would be delighted to go there.
  5. Semi-fixed or fixed phrases made up of common de-lexicalized words (words which carry little or no definite meaning when taken out of context), such as "take," "get," "make," "have." For example, You've got to . . .; get upset; take your time; take care; take it easy.

The study further showed the following features of chunk use by JLEs engaging in the communicative tasks:

  1. Repeated use of a limited number of chunks with discoursal and pragmatic functions (e.g. I think; of course).
  2. Limited use of common collocations and "vague language" which were frequently spotted in NS discourse.
  3. Sentences were mostly built "on the spot" on a word-by-word basis. This resulted in slow and far from fluent output. Moreover, some JLE discourse was unidiomatic and non-nativelike at times.
  4. Various instances of erroneous chunk use by the JLEs were found. For example, I don't have an idea (for "I have no idea"); As I told above (for "As I said above").

Although the results were not surprising, they lend support to previous claims (e.g. Pawley & Syder, 1983; Nattinger & DeCarrico, 1992; Lewis, 1993) that NSs use a lot of multiword chunks in their spoken discourse, emphasizing their important role in facilitating nativelike fluency and communicative competence.

The study also highlighted the fact that even JLEs with relatively high English proficiency lack the lexical competence to speak naturally and idiomatically. The limited chunk use by the JLEs indicated that they had to compose most sentences from scratch, thus slowing down language production time. More importantly, generating sentences from scratch at times led to unidiomatic or "non-nativelike" expressions.

As discussed above, the study showed that the NSs verbally realized some communicative strategies (e.g. giving evaluative responses) by uttering a wide variety of chunks, such as "That's interesting"; "That's a point"; "I liked your idea"; "That's true", etc. In contrast, the JLEs verbal realization of such a strategy was limited to repeated use of only a few expressions (most notably "I think so" and "I agree"), indicating that they might not have in their lexicons a wide enough repertoire of the multiword chunks commonly used by NSs.

These findings point to the fact that if learners are to become competent users of the English language, they should acquire and use multiword chunks in their discourse. It also becomes apparent that in EFL situations, such as in Japan, where most English language learners very rarely have the benefit of learning from direct interactions with NSs, one effective way to acquire multiword chunks, especially those with pragmatic functions, is to be exposed to authentic NS spoken discourse in the language classroom.

Pedagogical Implications

The above discussion underscores the importance of raising learner consciousness in the use of multiword chunks in the language classroom. It also alerts teachers to the fact that:

  1. Learners need to acquire a) a wide repertoire of multiword chunks, and b) the ability to make native-like selections in the use and language manipulation of such chunks. To do this, learners should be exposed to spontaneous NS discourse so that they can see how and when chunks are manipulated and used by NSs. This can be achieved by the use of authentic teaching materials. Additionally, teachers should "increase carefully-controlled teacher-talking-time" (Lewis, 1997, p.52), i.e. they should constantly repeat and recycle common collocations and expressions while talking to their students.
  2. Learners should be made aware of the close relationship and integration of grammar and lexis. They need to learn commonly used multiword chunks for fluency, and they also need to know and use grammar, which is regulative in function, to adapt a lexical chunk to a particular context (Nyyssönen, 1995).
  3. The extensive use of sentence frames also suggests that "it may be possible to teach some of what has usually been considered as grammar in terms of vocabulary" (Kennedy, 1990, p.216). For instance, the word "would" is traditionally taught as "the conditional." However, the various common patterns it forms with other words can be more easily taught lexically as fixed/semi-fixed sentence frames, without analyzing their internal structures, e.g. "Would you like " (offering); "I would rather . . ." (showing preference), etc.
  4. 1Vocabulary learning should not be limited to the learning of single-word definition and usage. Instead, learners should be taught the various aspects of a word such as its common collocations, and related grammatical patterns (Nation, 1994). In particular, the more de-lexicalized a word is, the more important it is to teach the word in different contexts, showing its most common collocations, and different usage, etc. (Lewis, 1997).
  5. The common use of discourse and interactive devices by NSs informs us that these are useful devices and not "linguistic crutches" or "empty fillers" as viewed by some teachers (Nattinger & DeCarrico, 1992), and therefore should be included in the language learning syllabus. It also implies that an "overall preoccupation with lexis at clause- or sentence-level" (McCarthy, 1984, p.14) is not the right approach to teaching vocabulary. Instead, McCarthy suggests that we should examine the use of chunks in relation to other lexical units "(a) above sentence-level, (b) across conversational turn-boundaries, and (c) within the broad framework of discourse organization."

Some Classroom Activities

Outlined below are suggestions for adapting classroom activities to incorporate teaching multiword chunks to language learners. While some of the activities may have been used by language teachers for a long time, the main focus and emphasis here is on "chunk" acquisition and usage, rather than out-of-context, single, space-bound words.

Discovering the use of chunks through discussion and L1/L2 comparison

One effective way to sensitize JLEs to the use of common chunks with discoursal or pragmatic functions is to have students compare their own discourse with NS discourse. This can be done by asking JLEs and NSs to do the same communicative tasks. Their discourse can then be recorded and transcribed for JLEs to compare. Alternatively, NS spoken discourse from movie scripts, TV talk show transcripts, etc. downloaded from the Internet can be presented to students who can then discuss how some common chunks are used and what their functions are. For example, the following is an excerpt of a transcript of two NSs doing a language task:

A: Why don't we each give an example and then . . .
B: Fine with me.
A: Okay, now, . . .

Students have to identify all the chunks therein and discuss their functions. (e.g. "Why don't we" is a sentence frame for making suggestions. "Fine with me" is a chunk with the pragmatic function of showing agreement. "Okay, now" is a chunk with the discoursal function of a shifting topic.) Students are also encouraged to suggest L1 equivalents.

As Yorio (1980, p.440) points out, the emphasis here should be placed on raising learners' awareness of "sociolinguistic variables, not on immediate production". This can be best achieved by discussion and L1/L2 comparison.

"Spot-A-Chunk" activity to sensitize JLEs to the concept of "chunking"

Many JLEs tend to translate word for word from L1 to L2 or vice versa, ignoring the fact that many multiword chunks cannot be analyzed and broken down to single words. Thus, giving students a short article or transcript of conversation and asking them to spot all the chunks therein is a good exercise for making them realize what constitutes a "chunk" and how frequently chunks are used in English. After spotting all the chunks, various activities can be done, for example, have students suggest L1 equivalents for some of the fixed idiomatic chunks. Some teachers may think that this is a reversion to the outdated "Grammar-translation" method. However, as Lewis (1997) points out, many multiword chunks, especially those with pragmatic functions, do have equivalents in other languages. Having students translate chunk-for-chunk (not word-for-word) has the advantage of getting students to associate chunks in L2 with their L1 equivalents, which can greatly reduce the burden of L2 learning on the students. This exercise is especially useful for spotting chunks formed by de-lexicalized words. Many students tend to ignore "small" words (e.g. "take," "get," "point") because they understand the literal meaning of such words. However, many useful chunks formed by such words are idiomatic, and their real meaning is hard to decipher, e.g. "take your time," "take it easy," "take a break," "take your point." Identifying and learning how to use these chunks will greatly increase the students' vocabulary size and fluency.

For sentence frames or chunks that can have different words in them (e.g. The thing/point/problem is . . .), students can be asked to suggest which word(s) can be substituted in the frame and the respective meanings of each.

Using corpora and concordances in the classroom

With the advance of computer technology, corpora and concordances are now available relatively easily to teachers for use in the language classroom. The use of corpora and concordances is advantageous because it encourages students to discover language use on their own (Tribble & Jones, 1990). Corpora for classroom use can be compiled using various materials depending on the needs and skill levels of students. Some good sources of materials include academic textbooks, movie scripts, news and talk show transcripts downloadable from various sites on the Internet, etc. (For detailed instructions on how to compile a corpus, please refer to John,1997.)

By running a corpus through a concordancing program (e.g. ConApp; Web Concordancer; WordSmith), a concordance list for a certain word can be generated (see Appendix B). Students can be asked to look at such a list and discover for themselves the most common uses of, for example, a de-lexicalized word and its collocations, and the contexts with in which such collocations are used. They can then share their discoveries with the whole class. Teachers can use concordance lists to create language exercises such as gap filling exercises. For ESP teachers, a corpus made up of a single theme or subject can be created, and used to show students the high frequency chunks in a specific content domain.

Short language exercises and games

Commonly used language exercises and games can easily be adapted to give students quick, short spurts of chunk practice. Below are just a few:

  1. Scrambled Sentences
    Students have to put the words in the right order. The sentences should be, or contain, common multiword chunks. For example:
    it/for/go - go for it
    its/or/never/now - it's now or never
  2. Hangman
    Use short fixed, commonly used expressions. The function of the expression can be given to students as a hint. For example: G O F O R I T (Hint: An expression used to encourage someone to try something)
  3. De-lexicalized Word Race
    Give the class a de-lexicalized word (e.g. take). In small groups, students have 5 minutes to come up with as many common expressions containing the word as possible. For example:
    take a bath; take your time; take it easy . . .

The above are just a few activities, among many, which are useful for raising learner awareness in L2 chunk use. More suggestions can be found in Lewis's Implementing the Lexical Approach (1997, Chapters 6 & 7).

Conclusion

The fact that native speakers of English use a lot of multiword chunks in their discourse is indisputable. Equally indisputable is the important role played by these chunks in the facilitation of communicative competence. However, it is still not common practice for language teachers to systematically introduce such chunks to L2 students. Although a few language textbooks have started to draw learners' attention to the concept and importance of chunks, the majority of the textbooks writers with books on the market have been slow to catch on. Before learners can be taught such chunks, a paradigm shift in the concept of "vocabulary teaching" has to be embraced by language teachers and material writers alike. This can only be achieved through more disciplined research in this area. Teachers are also encouraged to take on the role of researchers by investigating more into the natural language of English.

Acknowledgements

I would like to thank Jane Willis and the two anonymous reviewers for their invaluable comments and suggestions on earlier drafts of this paper. Thanks, also, to David Cox for allowing me to use his research corpora in my study.

Hazel Ketko received her MSc in TESOL from Aston University, Birmingham, U.K., and is now teaching EFL at Hiroshima International School and Hijiyama University. She is interested in corpus analysis and computational linguistics. <ketko@fureai-ch.ne.jp>

References

Aston, G. (1995). Corpora in language pedagogy: Matching theory and practice. In Cook, G., & Seidlhofer, B. (Eds.), Principle and practice in applied linguistics: Studies in honour of H.G. Widdowson. Oxford: Oxford University Press.

Bolinger, D. (1975). Aspects of language (2nd Edition). New York: Harcourt Brace Jovanovich.

Bolinger, D. (1976). Meaning and memory. Forum Linguisticum, 1(1), p. 1-14.

Carter, R. (1998). Vocabulary: Applied linguistics perspectives. London: Routledge.

Channell, J. (1994). Vague language. Oxford: Oxford University Press.

Cox, D. (1998). Can we predict language items generated by "open" tasks? Unpublished MSc dissertation. Birmingham: Aston University.

Johns, T. (1997). Improvising corpora for ELT: Quick-and-dirty ways of developing corpora for language teaching. Available at: <web.bham.ac.uk/johnstf/palc.htm>.

Kennedy, G.D. (1990). Collocations: Where grammar and vocabulary teaching meet. In Sarinee, A. (Ed.), Language teaching methodology for the nineties. ERIC Document No. ED366196.

Ketko, H. (2000). A comparative study in the use of multiword chunks between English native speakers and Japanese learners of English. Unpublished MSc dissertation. Birmingham: Aston University.

Lewis, M. (1993). The lexical approach. Hove: Language Teaching Publications.

Lewis, M. (1997). Implementing the lexical approach. Hove: Language Teaching Publications.

McCarthy, M. (1984). A new look at vocabulary in EFL. Applied Linguistics, 5(1), 12-22.

Moon, R. (1997). Vocabulary connections: multiword items in English. In Schmitt, N., & McCarthy, M. (Eds.), Vocabulary description, acquisition and pedagogy. Cambridge: Cambridge University Press.

Nation, I.S.P. (1994). New ways in teaching vocabulary. Alexandria, Virginia: TESOL, Inc.

Nattinger, J., & DeCarrico, J. (1992). Lexical phrases and language teaching. Oxford:

Oxford University Press.

Nyyssönen, H. (1995). Grammar and lexis in communicative competence. In Cook, G., & Seidlhofer, B. (Eds.), Principle and practice in applied linguistics: Studies in honour of H.G. Widdowson. Oxford: Oxford University Press.

Pawley, A., & Syder, F. (1983). Two puzzles for linguistic theory: nativelike selection and nativelike fluency. In Richards, J., & Schmidt, R. (Eds.), Language and communication. London: Longman.

Peters, A.M. (1983). The units of language acquisition. Cambridge: Cambridge University Press.

Skehan, P. (1995). Analysability, accessibility, and ability for use. In Cook, G., & Seidlhofer, B. (Eds.), Principle and practice in applied linguistics: Studies in honour of H.G. Widdowson. Oxford: Oxford University Press.

Tribble, C., & Jones, G. (1990). Concordances in the classroom: A resource book for teachers. London: Longman.

Widdowson, H.G. (1989). Knowledge of language and ability for use. Applied Linguistics, 10(2), 128-37.

Yorio, C. (1980). Conventionalized language forms and the development of communicative competence. TESOL Quarterly, 14, 433-42.

Appendix A

The communicative tasks used in the study

Task A

List the three most interesting cities or places in your country and say why people should visit them.

Discuss your ideas briefly and then say which of your partner's places you would most like to visit.

Task B

What advice would you give to the person who wrote this letter? Discuss your ideas and then agree on the two best suggestions.

Dear Angie:
My husband and I are worried about our daughter. She refuses to do anything we tell her to do and is very rude to us. Also, she has become very friendly with a girl we don't like. We don't trust her anymore because she is always lying to us. Are we pushing her away from us? We don't know what to do, and we're worried that she is going to get into trouble.

Worried Parents

Appendix B

First 12 Concordance lines for "take" (Brown Corpus)

 forget it all. But you just have to   take a few multivitamins and plough on.
 se next year when the school will   take a further cut because a reorganisat
 hought I would get a good degree,   take a law conversion course and become
 s must keep you busy - but please   take a moment to hear the views of an 'o
 But if moral theology is thus to   take account of biology, then it must su
 n asked: 'Is it practical? If you   take action against parents you tend to
 s minister, when asked if he would take   action on class size, 1991 W H A
 order or a personal computer can   take advantage of materials in an easy-t
 with Springer and Mark Fielding to   take advantage of tax changes that allow
 us fought to get this far that we   take 9am starts very seriously. We had
 was one of the first in Britain to   take an interest in this music, invitin
 living in TV and theatre, which I   take as success. But in answer to the q


All materials on this site are copyright © by JALT and their respective authors.
For more information on JALT, visit the JALT National Website