Proceedings of the 2nd Annual JALT Pan-SIG Conference.   May 10-11, 2003. Kyoto, Japan: Kyoto Institute of Technology.
Promoting fluency in EFL classrooms
by James Dean Brown    University of Hawai'i at Manoa   


This paper begins by exploring a number of definitions of fluency found in the literature. The paper then proposes that teachers should provide students with a complete set of language tools: (a) paralinguistic features, (b) kinesic language features, and (c) pragmatic tools, as well as expanded views of (d) pronunciation (to include suprasegmentals and reduced forms), (e) grammar (to include the grammar of spoken discourse), and (f) vocabulary (to include multiple word meanings, idioms, and informal vocabulary). In addition, teachers should help students learn to make language choices that have to do with (a) settings, (b) social, sexual, and psychological roles, as well as (c) register and style. Teachers can also help students develop six language strategies: the abilities to (a) use speed to their advantage, (b) use pauses and hesitations efficiently, (c) give appropriate feedback, (d) repair competently, (e) clarify effectively, and (f) negotiate for meaning when necessary. Finally, teachers can use five approaches in their classrooms that will promote fluency: (a) encourage students to go ahead and make constructive errors, (b) create many opportunities for students to practice, (c) create activities that force students to focus on getting a message across, (d) assess student's fluency not their accuracy, and (e) talk to the students about fluency.

Keywords: fluency, communicative language features, foreign language instruction, natural language features

a)場面、b) 社会的、性別、心理的な役割、c)レジスターとスタイル、といった言葉の選択の必要性を学習させる
   最後に、教室での流暢さの指導に関して教師が取る5つのアプローチについて提案する。それらは、 a)学生が


What is fluency? According to Hartmann and Stork (1976, p. 86), "A person is said to be a fluent speaker of a language when he can use its structures accurately whilst concentrating on content rather than form, using the units and patterns automatically at normal conversational speed when they are needed."
Fillmore (1979) proposed that fluency includes the abilities to:
  1. fill time with talk [i.e., to talk without awkward pauses for a relatively long time]
  2. talk in coherent, reasoned, and "semantically dense" sentences [Fillmore's emphasis]
  3. have appropriate things to say in a wide range of contexts
  4. be creative and imaginative in using the language
Notice that Fillmore's notion of fluency is limited to oral productive language. Since fluency may not be limited to oral productive language, I will have to come back to this issue.

[ p. 1 ]

Brumfit (1984, p. 56) felt that fluency meant, "to be regarded as natural language use." He also mentioned Fillmore's four fluency skills and pointed out that they are related to four characteristics: speed and continuity, coherence, context-sensitivity, and creativity. He further argued that the characteristics are related to four "basic sets of abilities": psycho-motor, cognitive, affective, and aesthetic (p. 54).
Richards, Platt, and Weber (1985, p. 108) define fluency as follows: "the features which give speech the qualities of being natural and normal, including native-like use of pausing, rhythm, intonation, stress, rate of speaking, and use of interjections and interruptions." They further point out that, in second and foreign language learning, fluency is used to characterize a person's level of communication proficiency, including the following abilities to:
  1. produce written and/or spoken language with ease
  2. speak with a good but not necessarily perfect command of intonation, vocabulary, and grammar
  3. communicate ideas effectively
  4. produce continuous speech without causing comprehension difficulties or a breakdown of communication. (Richards et al, 1985, pp. 108-109)
Lennon (1990, p. 388) says that the term fluency is used in two different ways in the literature, what he calls its broad and narrow meanings. The broad definition operates "as a cover term for oral proficiency," which "represents the highest point on a scale that measures spoken command of a foreign language" (p. 389). The narrow definition of fluency is that it is "one, presumably isolatable, component of oral proficiency. This sense is found particularly in procedures for grading oral examinations..." (p. 389).
Schmidt (1992) describes fluency as an automatic procedural skill (citing Carlson, Sullivan, & Schneider, 1989). According to him, L2 fluency is a performance phenomenon which "depends on procedural knowledge [citing Faerch and Kasper, 1984], or knowing how to do something, rather than declarative knowledge, or knowledge about something."
". . . I feel that fluency can best be understood, not in contrast to accuracy but rather as a complement to it."

Sometimes fluency is defined in contrast to accuracy, which according to Richards et al (1985, p. 109) "refers to the ability to produce grammatically correct sentences but may not include the ability to speak or write fluently" . Traditionally, accuracy has been taught not only in grammar, as suggested by Richards et al, but also in vocabulary and pronunciation. Of course, grammar, vocabulary, and pronunciation would be included in the notion of grammar if Richards et al were using the all-inclusive notion of big "G" Grammar. However, I am fairly certain that is not what they meant, and will therefore have to come back to the issue of big "G" Grammar in the context of fluency teaching. In general, however, I feel that fluency can best be understood, not in contrast to accuracy but rather as a complement to it. As Brumfit (1984) phrased it, "In no sense is it [accuracy] meant to imply that fluent language may not also be accurate language."
The above definitions provide a fairly good starting point because they include different perspectives on fluency and describe much of what fluency is. In my view, however, a full explanation of fluency must account for many other factors - factors that will be explored in this paper.

Enlarging our view of fluency

My classroom experiences promoting fluency in my students have convinced me that we can only help our students become fluent after we have enlarged and restructured our view of the components of language and our concept of what fluency means. In more detail (see Table 1), before students can ever have any chance at improving their fluency, teachers must expand their traditional boundaries of accuracy to offer rules of appropriacy including knowledge of the communicative language tools students must be able to use, the communicative language choices they should be able to make, and the communicative language strategies they must use to compensate for the fact that they, like all users of the language including native speakers, lack 100% knowledge of the language.

[ p. 2 ]

Table 1. An expanded view of language fluency.

paralinguistic features settings using speed to advantage
kinesic language features social roles using pauses and hesitations
pragmatics sexual roles giving appropriate feedback
pronunciation (expanded) psychological roles repairing competently
grammar (expanded) register clarifying effectively
vocabulary (expanded) style negotiating for meaning

Communicative Language Tools

Communicative language tools are the components learners need in order to actually use language. We must make available to our students all the language tools available to successful language users, not just a subset of those tools. In traditional classrooms, students are taught pronunciation, grammar, and vocabulary, but these three tools are not enough, nor are they broadly enough defined. Effective language users, whether native speakers or L2 learners, have a much wider range of tools available, of course including pronunciation, grammar, and vocabulary (broadly defined), but also including paralinguistic features, kinesic language features, and pragmatics.

Paralinguistics Features

Paralinguistic features are those features of the spoken language found outside the actual sounds being made. Some of the common paralinguistic features are facial expressions, head movements, hand gestures, eye movements, and eye gaze. Paralinguistic features are vital communication tools and all human beings (and some animals) employ them. In addition, these are the tools that many people fall back on when all other forms of communication break down, sometimes quite amusingly, especially in situations where we do not speak the language at all. Thus, effective use of these features will enhance fluency, and perhaps more important, these features may be the one set of tools that learners can fall back on when all other language tools have failed.

Kinesic Language Features

Kinesic language features have to do with the use of the body in communication. Thus I would label the communication facets of distance, touching, and posture as kinesic language features. These features are important because they can be used to communicate friendliness, concern, hostility, and many other complex emotions without complex pronunciation, grammar, and vocabulary. They are also important because the misuse of kinesics can cause serious misunderstandings. Over the years, the most salient example of miscommunication due to kinesics has occurred with my many male Arab speaking students who have often made me very uncomfortable by standing very close to me and even touching me in violation of the kinesic rules of my own language and cultural group, North Americans. It turns out that Americans are rather cold in this regard. Hence, students learning English need to learn such kinesic rules just as they need to learn the rules of grammar, not because their use of kinesics is wrong, but rather because of the ways it can lead to misunderstandings.


Pragmatics includes those facets of language that are directly related to the particular contexts and social situations in which the language is being used. In other words, pragmatics encompasses the relationships between real world knowledge (especially of social conventions) and the language being used in a specific context. Thus, the relationships between speaker and listener in that context would be one important aspect of pragmatics, including issues like power differences and social distance.
Pragmatics would be a simple issue if the pragmatic rules of all languages and cultures were exactly the same, but like all other aspects of language, considerable variation is found in the pragmatic rules of different languages and cultures. Since these differences can cause communication problems, they are differences that can and should be taught in order to improve students' fluency.

Big "G" Grammar: Pronunciation, Grammar, and Vocabulary Revisited

Sooner or later, we must also recognize that our traditional views of pronunciation, grammar, and vocabulary teaching are limiting our students' linguistic options. The language teaching we offer to our students should indeed include traditional pronunciation, grammar, and vocabulary, but only in their full versions (what I am calling big "G" Grammar) if we want to offer students the full range of these three sets of tools.


For example, let's consider pronunciation. Traditionally, we focus on phonemes (that is, the smallest units of distinctive sound) to the exclusion of suprasegmentals, reduced forms, and other crucial generalizations about the pronunciation of real spoken English. In effect we are only giving our students the basic facts and keeping them ignorant of other curcial information.

[ p. 3 ]

Suprasegmentals are aspects of pronunciation beyond the level of individual phonemes. Typically, suprasegmentals involve two or more phonemes. Those aspects of pronunciation often labeled the suprasegmentals of English are stress, intonation, and voice quality. These suprasegmentals are crucial for expressing ourselves accurately and for being understood by others. To illustrate, consider the fact that an utterance like "You are looking very nice tonight" can be said in a bored manner, sincerely, sarcastically, suggestively, sexily, and probably many other ways depending on how stress, intonation, and voice quality are used. In other words, the same pronunciation, grammar, and vocabulary can mean a variety of things depending on stress, intonation, and voice quality - the suprasegmentals.
Reduced forms were defined in Brown and Hilferty (1989) as "native speakers' connected speech replete with its contracted forms, elision, liaison, and reduction." At this point in time, I would add several other subcategories: weak forms, linking, assimilation, and intrusion. Many people think of reduced forms as signs of lazy, slovenly, or careless English, but that view is naive and incorrect. Indeed, there is ample evidence that reduced forms are used (sometimes to lesser or greater degrees) in all kinds of English, even the most careful and formal.
Why is it that we do not teach these reduced forms? Instead many teachers seem to prefer to cover the written language and assume that students will automatically be able to understand the spoken language. But spoken English is very different from written English. Why then are such teachers amazed when their students cannot even understand natural spoken language like whenduyawannagonnagedoudahere? [When do you want to get out of here?], much less produce it. (For more on this relatively new and interesting topic, see Henrichsen, 1984; Brown and Hilferty, 1986a & b, 1987, 1989, 1995, & 1998; Weinstein, 1982, 2001; Ito, 2001; Rosa, 2002).
Crucial generalizations about real spoken English are also forgotten in the rush to teach phonemes. Some of those generalizations are as follows: (a) unstressed vowels in English become schwa in many environments, which makes schwa the most common (often ignored) vowel in American English, (b) phonemes change depending on the phonemes that come before and after them, (c) correct word stress and timing are crucial to being understood, and (d) speakers of syllable-timed (or mora-timed) languages often have trouble speaking English because English is largely stress timed.


Our view of grammar may be similarly antiquated – a fact that may limit our students' options. Consider the notion of requiring students to use complete sentences when they speak. While requiring complete sentences may be useful in teaching formal writing, it is not appropriate for teaching speaking. That is why we use the notion of utterances rather than sentences in describing spoken discourse. The fact of the matter (unrecognized by many) is that spoken discourse does not require complete sentences. Indeed, the persistent use of complete sentences will sound strange. Consider the following example conversation between a native speaker (NS), who uses short utterances, and a non-native speaker (NNS), who uses complete sentences just as she was taught:

	NS:   Howzit goin'? 
	NNS:  I am doing very well, thank you.
	NS:   Your family?  
	NNS:  They are doing very well, too, thank you.
	NS:   Whereya going?
	NNS:  I am walking to the snack bar. Would you like to come with me?

While she can be understood, it would be much more natural (and fluent) for the rules of conversational grammar and pronunciation to be applied instead of the rules of the written language, as follows:

	NS:   Howzit goin'? 
	NNS:  Fine, thanks.
	NS:   Your family?
	NNS:  They're good, too. 
	NS:   Whereya goin'?
	NNS:  The snack bar. Ya wanna come along?

Notice that the second conversation uses the grammar of spoken colloquial English, which does not impose the use of complete sentences, but rather uses utterances. Also note that several of these utterances appear to be chunks of language (e.g., "Howzit goin'?" and "Fine, thanks."), which would probably be better taught as units rather than as sentences assembled on the basis of grammar rules. Finally, notice that the pronunciation in the second example is reduced and much more appropriate.


Our view of vocabulary may also be old fashioned — a fact that could be limiting our students' options. Why is it that we let students learn one meaning for each word, when we know for a fact that words have many meanings? And, why is it that we avoid idioms? Truly fluent speakers must have command of the vocabulary and idioms of the language. Also, why is it that we never teach informal vocabulary like "vulgar" expressions? These types of vocabulary are part of the language (see Claire, 1990), and for students (men and women alike) to function in some of the most relaxed and informal language styles fluently, they must have a command of this vocabulary, too, at least receptive command.
Big "G" grammar includes many components. I remember learning in the 1970s that big "G" grammar should include pronunciation, little "g" grammar, and vocabulary. At this point in my career, I truly believe that each of those categories should be expanded and that additional major categories should be added. Pronunciation should be expanded to include phonemes, as well as suprasegmentals, reduced forms, and other useful generalizations about pronunciation. Grammar should be expanded to include the rules of written sentences and the rules of utterances in spoken discourse. Vocabulary should be expanded to include traditional word lists, as well as at least passive knowledge of multiple word meanings, idioms, and vulgar expressions. Additional major categories that need to be taught are paralinguistic features, kinesic language features, and pragmatics. We need to provide our students with the full range of tools they will need if they are ever to become fluent in English.

[ p. 4 ]

". . . I believe ESL/EFL curriculum is typically very inefficient, that we teach students the same things over and over again."

The arguments that I often get when I make these observations publicly are that (a) there just isn't enough time to cover all these items and (b) the students will pick up these things anyway after they get to the country.
As for the first argument, that there just isn't enough time, I believe ESL/EFL curriculum is typically very inefficient, that we teach students the same things over and over again. This happens because we never do diagnostic testing at the beginning of our courses to see what the students already know. We cover the same points over and over again, and then we wonder why the students are bored. But since the students do well on the final examination, we figure that they learned a lot. The fact that they could perhaps have done well on that final examination on the first day of class is beside the point, right? There just isn't enough time? Time could be made and we could do it by being better organized and more efficient.
As for the second answer, that students will pick it up when they get to the country, one could argue that students will "pick up" all aspects of the language once they get to the foreign country if given enough time, so why teach language at all? That's ridiculous, right? Unfortunately, the "they will pick it up" argument is saying that we should send partially informed humans into these culturally trying situations? People with only three out of the six sets of language tools available to them are communicatively handicapped, especially given that they have incomplete views of even those three sets of tools. To me, the purpose of foreign language teaching is to make the language learning process more efficient and less painful, and to prepare students for real-life communication. If that is the purpose, how can we defend sending students into real-life communication without the full range of tools they will need in their big "G" Grammar in order to be fluent and effective communicators.
From a practical day-to-day teaching point of view, the most important benefit of this expanded view of language tools for hardworking teachers is that students will find these topics interesting. Here is a challenge – try some of these topics and see if students don't find them interesting:
  1. Differences between L1 and L2 in how facial expressions, eye movements, eye gaze, head movements, and hand gestures are used
  2. Differences between L1 and L2 in distance, touching, and posture during conversation
  3. Differences between L1 and L2 in the pragmatic rules of conversation in different contexts and situations, as well as the rules related to the relationship(s) between speaker and listener
  4. Differences between L1 and L2 in how stress, intonation, and voice quality are used to affect meaning
  5. The importance of reduced forms in comprehending native English speech
  6. The importance of schwa in English pronunciation
  7. The importance of word stress and syllable timing to native speaker comprehension of non-native speech
  8. The importance of using and understanding utterances (rather than complete sentences) in spoken English
  9. The importance of learning some aspects of English as chunks rather than as vocabulary items and grammar rules
  10. The multiple meanings of words
  11. Idioms as important elements of vocabulary in English
  12. Informal vocabulary like swear words and vulgarisms in English
If I am right, covering some or all of these topics will not only provide students with tools they will need to increase their fluency in English, but will also make the classes more interesting for the students, in fact, much more interesting than the standard triad of traditional pronunciation, grammar, and vocabulary.

Communicative Language Choices

Communicative language choices are alternatives within the sets of language tools that students will need to select due to the context in which communication is taking place. Widdowson (1978, p. 13) made a distinction between what he called reference rules and expression rules. Reference rules are those rules that make up the student's knowledge of the language. Given the earlier discussion, a student's reference rules would consist of what they know of the paralinguistic, kinesic, pragmatic, pronunciation, grammar, and vocabulary rules of English. The expression rules would be those rules that determine what the student actually does with the language. Thus the choices that students will make when they actually communicate in English must be based on expression rules. These expression rules frequently center on the choices they must make due to differences in (a) settings, (b) social, sexual, and psychological roles, and (c) register and style.


Obviously, the setting is the place where the communication takes place. For instance, communication can take place (among many other places) on a cell phone, in front of a classroom, at a gym, at the auto mechanic's garage, in a dentist's chair, at a grocery store, in a library, in a movie house, in front of a large meeting of 250 people, in front of a small meeting of seven people, etc. Imagine for a moment the differences in language that would be necessary in those various settings. I would expect differences in paralinguistics, kinesics, pragmatics, pronunciation, grammar, and vocabulary, and I would not be disappointed. To provide just a few extreme examples, a person talking on a cell phone will generally speak louder and more effusively than a person speaking in a library. Or, a person who is lecturing to a large university class is likely to speak quite differently from a person working with a small kindergarten class.

[ p. 5 ]

Sexual, Psychological, and Social Roles

Within various situations, sexual, psychological, and social roles may further determine the choices that people make in communicating.
Sexual roles. The first set of roles, sexual roles, involves differences in the ways females and males communicate with other females and males. The research to date indicates that females and males behave differently in terms of the amount they talk in adult interactions. "Frequently men talk more than women; however, they do not necessarily do so . . . these behaviors are best explained in terms of the social structure of the interaction; this is informed by the difference in men's and women's abilities and areas of competence" (James and Drakich, 1993, p. 301). Thus learners may have to take into account sex role differences in amount of talk when making language choices, but also differences in interruption strategies (James and Clarke, 1993; Tannen, 1994) and coherence (Tannen, 1994), among other things. [For much more on gender differences in communication see Tannen, 1993, 1994.]
Psychological roles. The second set of roles, psychological roles, involves differences based on personality, aggressiveness, dominance, size, etc. For instance, small, passive persons are likely to speak differently to large, aggressive persons that they do to other small, passive persons. Thus various language choices may be based on psychological roles - role differences that students need to understand, especially if those roles are different in their culture.
Social roles. The final set of roles involves differences in the social roles that people play in life and how those roles affect communication. We all take on a variety of social roles. For instance, in my personal life, I am a father, son, uncle, husband, friend, ex-husband, etc. In my work, I am a friend, colleague, professor, administrator, employee, committee chair, committee member, etc. In my daily life, I am also a customer at a bank, grocery store, auto garage, drug store, dentist's office, doctor's office, etc. In some roles, I am humble and meek, for instance, when I take my car in for servicing, I know absolutely nothing about automobile mechanics and feel rather powerless, so my linguistic choices are those of a humble, meek (and perhaps stupid) customer. In other roles, it is quite natural for other people (students, for instance) to be humble and meek around me. Clearly then, we are making language choices throughout our days as we move from social role to social role.

Register and Style

Over the years, some confusion has arisen over the definitions and use of the terms register and style.
Register. Register refers to differences in language choices based on membership in different occupations (e.g., professors, students, mechanics, sportscasters, ministers, etc.) or different hobbies or interest areas (e.g., skating, skiing, model railroading, computer hacking, fishing, etc.). Membership in each of these groups undoubtedly involves the use of specialized vocabulary, and may even include changes in grammar and pronunciation. For example, a protestant minister typically uses a register that includes considerable specialized vocabulary (e.g., Baptism, Host, Holy Trinity, pulpit, salvation, sermon, etc.), will often use specialized grammar (e.g., thy will be done, hallowed be thy name, etc.), and will sometimes use particularly stentorious pronunciation while delivering a sermon.
Style. Style describes differences in level of formality, differences that range from very casual to casual to colloquial to formal to very formal. Such differences in style are usually made in response to differences in settings, differences in social, sexual, and psychological roles, and differences in register. However, style can vary within any of these categories, as well, based on degrees of personal relationship and other factors. For example, a minister might be very formal at the pulpit during his sermon and suddenly change to more colloquial while greeting parishioners after the sermon and become very casual while greeting a particularly close friend - all different styles within a minister's register.

Making Correct Communicative Language Choices

The ability to switch registers and styles and respond appropriately to different sexual, psychological, and social roles in various settings is one mark of a fluent English speaker. Yet often, we give our students no information about the various expression rules involved in these choices. Naturally, such rules will become increasingly important as they become relatively advanced in the language (or as they get closer to leaving for the country). However, that does not mean we cannot start teaching them some aspects of these expression rules from the very beginning. For example, if you teach your students "How do you do?" at an early stage, it might be wise to warn them that this is a very formal expression seldom used by Americans (except in very formal situations) and that "How are you?" or "Pleased to meet you" are less formal and usually more appropriate ways to say the same thing. Instead of giving them one usually inappropriate language option, give them several options so they can begin to make their own language choices.

[ p. 6 ]

Communicative Language Strategies

Communicative language strategies are abilities that students need in order to maximize communication when they are less than 100 percent accurate in their use of language. Fluency is probably not an absolute characteristic that students either have or do not have. If, in fact, fluency is a matter of degrees, students at any level of proficiency can probably achieve some degree of fluency. Communicative language strategies can help them communicate fluently with whatever proficiency they happen to have at any given time. There are at least six such strategies, including the abilities: (a) to use speed to their advantage, (b) to use pauses and hesitations efficiently, (c) to give appropriate feedback, (d) to repair competently, (e) to clarify effectively, and (f) to negotiate for meaning when necessary.

Ability to Use Speed to Their Advantage

Untrained teachers will often think of fluency as being about speed. But fast speech is not necessarily fluent speech. In fact, fluent native speakers will vary their speed depending on the context in which they are speaking. My experience is that non-native speakers of English also think that they should speak fast to be more fluent. However, fast speech is not automatically fluent speech. That is a message that students need to understand, i.e., it is fine to speak relatively slowly as long as it is done at an appropriate speed. Indeed, native speakers often speak rather slowly in order to have time to think as they talk. Hence, the appropriate speed is the speed at which speakers can think clearly and still succeed in getting their message across.

Ability to Use Pauses and Hesitations

Native speakers of English often use pauses and hesitations while they are talking. Indeed, researchers have found that native speakers spend up to 50 percent of their speaking time pausing. (For more on pauses, see Clark and Clark, 1977; Hatch, 1983; or Tarter, 1986). Learners need to understand this aspect of communication. However, students will often refuse to believe that natives pause and hesitate, so teachers may need to audiotape or videotape some actual native speakers talking in a natural situation (i.e., not actors) in order to demonstrate to students that natives actually do hesitate and pause, and they do so quite often. One way or the other, students need to understand that pauses and hesitations are necessary and natural parts of spoken language. The reason for all this pausing and hesitating is that humans need time to think when they are talking, and they use pausing and hesitating to give themselves that time.
However, pauses and hesitations are not just dead time. Students need to understand that using slower speed with pauses and hesitations may necessitate the use of fillers. Some fillers in English are just sounds like uhm, er, uh, ah, and umm; other fillers are words such as okay, you know, well, so, etc. The purpose of such fillers is to fill silence, which in turn makes communication seem more natural – and fluent. Native speakers of English do hesitate and pause, but they do not seem to be able to tolerate long silences. Instead, they will use fillers to appropriately avoid long silences. A nice side effect of all this is that the process of using fillers also gives them time to think.

Ability to Give Appropriate Feedback

Feedback consists of all the signals that a listener directs at a speaker to indicate that the message is (or is not) getting through. Feedback can express agreement or disagreement, understanding or misunderstanding, comprehension or confusion, etc. In addition, the signals used to express these meanings can include not only sounds and words, but also gestures and facial expressions. Sounds could include grunts of agreement, sounds like mm, uh huh, hmm, etc. Words might include feedback signals like really, yep, right, yeah, okay, etc. Gestures might include a hand signal to continue or stop, or talk faster, or head signals like nodding agreement, showing wonderment, etc. Facial expressions could involve smiles, winks, frowns, quizzical expressions, direct eye contact, evasive eye contact, etc. Teachers should not simply side step this set of strategies by saying that students will "pick them up" easily along the way. Such feedback signals should be taught because they are clear and obvious indicators of fluency that can make a person seem very foreign or very fluent depending on how well they are used.

Ability to Repair Competently

Native speakers of English are sometimes guilty of mispronunciation, false starts, back tracking, stuttering, etc., but in most cases, they feel nothing similar to the embarrassment that many non-native speakers feel when they make such "mistakes." One difference between these two groups of people is that the native speakers know how to handle their mistakes better, that is, how to competently repair them. When speakers detect flubs (I think that is the technical term) of various kinds and correct themselves, it is called self-repair. When one speaker corrects another, it is called other-repair. In order to avail themselves of this strategy, students need to be taught how to correct their own errors, how to understand and accept corrections from others, and eventually, how to correct errors that others make without creating offense.

[ p. 7 ]

Ability to Clarify Effectively

Clarification is one type of repair that is particularly important. When a fluent speaker spots that a listener is not understanding (through verbal signals, gestures, facial expressions, etc.), the speaker will typically try to clarify by rephrasing, defining terms, summarizing, using gestures, drawing a picture, etc. The point is that fluent speakers, when misunderstood, will use whatever strategies they can to clarify their message.

Ability to Negotiate for Meaning When Necessary

Likewise, fluent speakers who are failing to understand something in a conversation will use whatever verbal signals, gestures, or facial expressions are necessary to get the other speaker to clarify. And if they still do not understand after the speaker clarifies, fluent speakers will again seek clarification. This process of give and take in conversation (including various interactions of feedback, repair, and clarification) is called negotiation. Typically, negotiation is focused on cooperating to get meaning across; in the process, it sometimes centers on clarifying vocabulary, grammar, or even pronunciation details.

Promoting Fluency

". . . teaching fluency is different from teaching other aspects of language. In teaching fluency, we must be willing to let go of some of the control in our classrooms"

Can we really teach fluency? I think so, but we may have to modify the traditional ways in which we conceive of teaching. We can certainly teach fluency by giving lectures that help expand our students' knowledge of the choices, tools, and strategies at their disposal. However at a certain point, we will have to admit that teaching fluency is different from teaching other aspects of language. In teaching fluency, we must be willing to let go of some of the control in our classrooms; we must be willing to let the students have some of the control and let them do some of the work; we must be willing to set up situations in which fluency can develop, and then encourage the students to actually communicate. I'm not saying that we need to teach fluency all of the time, but I am saying that some of the time students need a little guided communication time during which their knowledge of the many aspects of the language can develop into fluency.
Unlike language knowledge, fluency is about automatizing the language knowledge. As Schmidt (1992) said, "Fluent speech is automatic, not requiring much attention, and is characterized by the fact that the psycholinguistic processes of speech planning and speech production are functioning easily and efficiently." Such automaticity can only occur when the students themselves are trying to use their language knowledge to actually communicate, and we can only help the students become fluent by creating opportunities for them to practice communicating and then stepping out of the way (for example, see the principles set out by Gatbonton and Segalowitz, 1988).
Having taught speaking in China and elsewhere, I have learned that teachers can promote the ease and efficiency associated with automaticity in speech production. As I explained at some length in Brown (1996), teachers can promote fluency if they: (a) encourage students to go ahead and make constructive errors, (b) create many opportunities for students to practice, (c) create activities that force students to get a message across, (d) assess student's fluency not their accuracy, and (e) talk openly to the students about fluency.

Encourage Students to Go Ahead and Make Constructive Errors

Many students ferociously concentrate on producing accurate grammar. In doing so, they may be obviating the possibility of ever becoming fluent. Especially, in some of the Asian countries that I have worked in, students are afraid of making mistakes because they do not want to "lose face" in front of their peers. Making errors is therefore a topic that I have had to teach to my students. I have used three strategies: (a) explain native speaker error patterns, (b) minimize error correction, and (c) treat error making as a skill that students must master. [For an excellent in-depth book on the topic of errors, see James, 1998.]
As I explained in more detail in Brown (1996), one key to encouraging students to make constructive errors is to explain native speaker error patterns, which involves at least four steps. First, tell the students that native speakers of English make errors in pronunciation, word choice, grammar, even logic. Second, since learners may not be close enough observers to recognize that native speakers make errors, you may have to illustrate that fact by pointing out errors in your own speech (since teachers' speech in class is often highly monitored, you may need to ad lib a few flubs) or in tapes of other native speakers. Third, do whatever is necessary to help students recognize that it is okay for them to make production errors, indeed, it is a natural part of all communication, even among native speakers. Fourth, tell them that, if they remain unwilling to make errors, they will probably never be able to become fluent, i.e., that sometimes they need to focus on accuracy but other times they have to relax and practice the automaticity that is necessary for fluency development.
Another key to encouraging students to make constructive errors is to minimize error correction. If teachers seriously want to foster fluency, they absolutely must limit error correction to those errors that hinder communication. In a sense, the student's responsibility in fluency development is to bring the level of their English production (especially in speaking) up to the level of their knowledge of grammar. "Fluency, then, can be seen as the maximally effective operation of the language system so far acquired by the student" (Brumfit, 1984, p. 57). During periods of fluency development, teachers should therefore avoid yanking the students back to a focus on accuracy (which is what happens if they correct trivial errors) because that might bring the whole fluency development process to a halt.

[ p. 8 ]

A third key to encouraging students to make constructive errors is to treat error making as a skill. First, students need to understand that they can only learn fluency by making errors and learning how to deal with those errors. Second, they need to understand that errors are a natural component of language development, not an indication of their lack of worth as human beings. Third and most important, they need to develop a willingness to make errors. As I put it in Brown (1996), "A student who is afraid to make errors won't make errors, and a student who won't make errors won't become fluent. For many students, this may mean learning to take chances in ways that they have never done before."

Create Many Opportunities for Students to Practice

In addition to encouraging students to make constructive errors, teachers should provide ample opportunities for students to practice fluency development. This means creating speaking courses, wherein teachers set up activities and then get out of the way so that many students can be talking at the same time (in pairs or small groups). We should avoid at all costs conversation courses, wherein the teacher does much of the talking and students respond to the teacher one at a time. For the many teachers who were educated in teacher-centered classes, it is often difficult to set up student-centered activities like pair work, group work, role plays, etc. and then simply let the students go. However, setting up such activities is exactly what the students need to develop their fluency. I also find that creating a relaxed classroom atmosphere (by using humor, songs, personal interactions, smiles, cartoons, etc.) helps promote fluency.

Create Activities That Force Students to Get a Message Across

Nowadays, teachers have numerous resources to fall back on in creating communicative activities (for instance, see Keller and Warner, 1979; Sadow, 1982; Klippel, 1987; Fried-Booth, 1988; Ladousse, 1988; Bailey and Savage, 1994). Whether selecting fluency activities from sources like those just listed, or creating activities for a specific situation, teachers should insure that all fluency activities focus the students' attention on getting their meaning across. A meaning focus can be achieved by creating activities wherein students have a specific task to perform, a particular problem to solve, or a clearly defined goal to reach. If the activities are properly designed, the students will become so involved in succeeding at their tasks that they will necessarily be focused on getting their meaning(s) across.

Assess Student's Fluency Not Their Accuracy

Whether we like it or not, students are very test driven. If we test them using multiple-choice grammar tests, they will prepare for multiple-choice grammar tests (and wonder why on earth we are doing pair work in class). If, in contrast, we test them in role plays, pair work, interviews, etc., the students will prepare for such assessment activities. Students are not stupid. They will prepare for the test we give them, and we must channel that energy so that they practice what we want them to practice. Mendelsohn (1992) provides a set of observation/scoring criteria that teachers might consider adapting for their own purposes. Mendelsohn gives guidelines for giving feedback on group activities as well as guidelines for feedback on presentations in front of the class. The categories of feedback he suggests are all focused on communicative aspects of language production (like turn taking, interrupting, etc.) or big "G" grammar topics (like those discussed above for paralinguistic features). Tellingly, he leaves issues of grammatical or phonological accuracy to the end and simply provides blank spaces for the teacher to write observations.

Talk Openly to the Students about Fluency

Unfortunately, students do not always readily agree with teaching methods like those I described above. For example, when I was teaching in China, my students protested: (a) that they couldn't learn from other students in pairs, groups, and such, and (b) that they preferred teachers who lecture on the finer points of grammar in English. They suggested that we (the foreign teachers) should watch how Chinese professors teach English if we want to see how it should be done. Clearly, convincing the students that our way of teaching was useful and important to them became a large part of our job:
  1. We pointed out to them that they had very high grammar test scores, but low scores in reading, writing, listening, and speaking. So, we argued, they should stop worrying about grammatical accuracy and turn instead to developing the ability to use their knowledge of the language for something useful, like speaking.
  2. We explained what we were trying to accomplish in developing their abilities to use speed to their advantage, to use pauses and hesitations efficiently, to give appropriate feedback, to repair competently, to clarify effectively, and to negotiate for meaning.
  3. We also shared with them why we were encouraging them to go ahead and make constructive errors, creating many opportunities for them to practice, creating activities that forced them to focus on getting a message across, assessing their fluency not their accuracy, and why we were talking to them about fluency.
  4. We asked for their feedback on the sets of ideas in 1, 2, and 3 above.
In short, we were honest with them about our intentions, we showed respect for their ability to understand what we were trying to do, and we solicited their feedback. These strategies of talking openly with the students helped convince them to go along with our "strange" teaching methods.

[ p. 9 ]


This paper has argued that students must learn three new sets of communicative language tools that are seldom taught in second or foreign language classes: In addition, students need to learn expanded versions of the traditional linguistic tools:
Students also need to learn how to make communicative language choices based on the following three sets of issues: In addition, students need to learn six communicative language strategies that they can use to increase their fluency, which are the abilities to: Finally, teachers can promote fluency by doing five things in their classrooms: One last point I would like to emphasize is that fluency is not an absolute issue that students either have or do not have. Instead, I would argue that fluency is a relative issue, even for native speakers who also vary in their fluency. Given that fluency is a matter of degrees, some degree of fluency can probably be achieved at all levels of language proficiency. Why not start today helping your students to develop fluency in English and offering them the full range of communicative language tools, choices, and strategies?


Bailey, K. M. & Savage, L. (Eds.). (1994). New ways in teaching speaking. Washington, DC: TESOL.

Brown, J. D. (1996). Fluency development. In G. van Troyer (Ed.) JALT '95: Curriculum and evaluation (pp. 174-179). Tokyo: Japan Association for Language Teaching.

Brown, J. D., & Hilferty, A. G. (1986a). Listening for reduced forms. TESOL Quarterly, 20(4), 759-763.

[ p. 10 ]

Brown, J. D. & A. Hilferty. (1986b). The effectiveness of teaching reduced forms for listening comprehension. RELC Journal, 17 (2), 59-70.

Brown, J. D., & Hilferty, A. G. (1987). The effectiveness of teaching reduced forms for listening comprehension. RELC Journal, 17 (2), 59-70.

Brown, J. D., & Hilferty, A. G. (1989, January). Teaching reduced forms. Modern English Teaching, 26-28.

Brown, J. D., & Hilferty, A. G. (1995). Understanding reduced forms. In D. Nunan (Ed.) New Ways in Teaching Listening (pp. 124-127). Washington, DC: TESOL.

Brown, J. D., & Hilferty, A. G. (1998). Reduced-forms dictations. In J. D. Brown (Ed.) New Ways of Classroom Assessment. Washington, DC: TESOL.

Brumfit, C. (1984). Communicative methodology in language teaching: The roles of fluency and accuracy. Cambridge: Cambridge University.

Carlson, R. A., Sullivan, M., & Schneider, W. (1989). Practice and working memory effects in building procedural skill. Journal of Experimental Psychology: Learning, memory, and cognition, 15, 517-526.

Claire, E. (1990). An indispensable guide to dangerous English for language learners and others (2nd ed.). Dundee, IL: Delta Systems.

Clark, H. H., & Clark, E. (1977). Psychology and language. New York: Harcourt, Brace, Jovanovich.

Faerch, C. & Kasper, G. (1984). Pragmatic knowledge: Rules and procedures. Applied Linguistics, 15, 214-225.

Fillmore, C. J. (1979). On fluency. In C. J. Fillmore, D. Kempler, & W. S. Y. Wang (Eds.), Individual differences in language ability and language behavior (pp. 85-102). New York: Academic Press.

Fried-Booth, D. (1988). Project work. Oxford: Oxford University.

Gatbonton, E., & Segalowitz, N. (1988). Creative automatization: Principles for promoting fluency within a communicative framework. TESOL Quarterly, 22, 473-492.

Hartmann, R. R. K., & Stork, F. C. (1976). Dictionary of language and linguistics. New York: Wiley.

Hatch, E. M. (1983). Psycholinguistics: A second language perspective. Rowley, MA: Newbury House.

Henrichsen, L. E. (1984). Sandhi-variation: A filter of input for learners of ESL. Language Learning, 34, 103-126.

Ito, Y. (2001). Effect of reduced forms on ESL learners' input-intake process. Second Language Studies, 20 (1), 99-124.

James, C. (1998). Errors in language learning and use. London: Longman.

James, D., & Clarke, S. (1993). Women, men, and interruptions: A critical review. In D. Tannen (Ed.) (1993). Gender and conversational interaction (pp. 231-280). Oxford: Oxford University.

James, D., & Drakich, J. (1993). Understanding gender differences in amount of Talk: A critical review of research. In D. Tannen (Ed.) (1993). Gender and conversational interaction (pp. 281-312). Oxford: Oxford University.

Keller, E. & Warner, S. (1979). Gambits conversational tools (Books one, two, & three). Hull, Quebec: Canadian Government Printing Office.

Klippel, F. (1987). Keep talking: Communicative fluency activities for language teaching. Cambridge: Cambridge University.

Ladousse, G. P. (1988). Role play. Oxford: Oxford University.

Lennon, P. (1990). Investigating fluency in EFL: A qualitative approach. Language Learning, 40 (3), 387-417.

Mendelsohn, D. (1992). Instruments for feedback in oral communication. TESOL Journal, 1 (2), 25-30.

[ p. 11 ]

Richards, J. C., Platt, J., & Weber, H. (1985). Longman dictionary of applied linguistics. London: Longman.

Rosa, M. (2002). Don'cha Know? A survey of ESL teachers' perspectives on reduced forms instruction. Second Language Studies, 21(1), 49-78.

Sadow, S. A. (1982). Idea bank: Creative activities for the language class. Rowley, MA: Newbury House.

Schmidt, R. (1992). Psychological mechanisms underlying language fluency. Studies in Second Language Acquisition, 14, 357-385.

Tannen, D. (Ed.) (1993). Gender and conversational interaction. Oxford: Oxford University.

Tannen, D. (1994). Gender and discourse. Oxford: Oxford University.

Tarter, V. C. (1986). Language processes. New York: Holt, Rinehart, and Winston.

Weinstein, N. (1982). Whaddaya say? Culver City, CA: ELS Publications.

Weinstein, N. (2001). Whaddaya say? Guided practice in relaxed speech (2nd ed.). White Plains, NY: Addison Wesley Longman.

Widdowson, H. G. (1978). The significance of simplification. Studies in Second Language Acquisition, 1, 11-20.

This paper is complementary to a paper presented at the JALT National Conference, Nagoya, Japan, 1995, which was later published in the proceedings for that conference as Brown (1996). That is, this paper expands on parts of the earlier paper that were given only brief coverage.

2003 Pan SIG-Proceedings: Topic Index Author Index Page Index Title Index Main Index
Complete Pan SIG-Proceedings: Topic Index Author Index Page Index Title Index Main Index

[ p. 12 ]