Teaching Grammar

DIANE LARSEN-FREEMAN

In "Teaching Grammar," Larsen-Freeman challenges conventional views of grammar. Instead of simply

analyzing grammatical form, she includes grammatical meaning and use as well. Then, building on
what is known about the way grammar is learned, she offers ways to teach grammar consistent with
contemporary theory and the need to "focus on form" within a meaning-based or communicative
approach.

INTRODUCTION
Over the centuries, second language educators
have alternated between two types of approaches
to language teaching: those that focus on analyzing the language and those that focus on using
the language. The former have students learn
the elements of language (e.g., sounds, structures, vocabulary), building toward students'
being able to use the elements to communicate.
The latter encourage students to use the language from the start, however falteringly, in
order to acquire it. Early in the previous century, this distinctive pattern was observable in the
shift from the more form-oriented grammartranslation approach to the use-oriented direct
method (Celce-Murcia 1980). A more recent
example of the shift is the loss of popularity of
the cognitive-code approach, in which analyzing
structures and applying rules are common practices, and the rise of more communicative
approaches, which emphasize language use over
rules of language usage (Widdowson 1978).
Even though such language use approaches
as task-based and content-based are in favor these
days, educators agree that speaking and writing
accurately is part of communicative competence,
just as is being able to get one's meaning across
in an appropriate manner. Further, it has been
observed that although some learners can "pick
up" accurate linguistic form from exposure to
the target language, few learners are capable of
doing so efficiently, especially if they are postpubescent or if their exposure is limited to the class-

room, as is the case when English is taught as a

foreign language. In contrast, research has shown
that teachers who focus students' attention on linguistic form during communicative interactions
are more effective than those who never focus on
form or who only do so in decontextualized grammar lessons (Spada and Lightbown 1993;
Lightbown 1998). It follows, then, that most educators concur with the need to teach grammatical
form. However, they advise doing so by "focusing
on form" within a meaning-based or communicative approach in order to avoid a return to analytic approaches in which decontextualized
language forms were the object of study.
Focusing on grammatical form during communicative interactions rather than forms in isolation (Long 1991) is one way to prevent the
pendulum from swinging beyond its point of
equilibrium. In this chapter, we will encourage a
balance between grammar and communication.
The first step is to come to a broader understanding of grammar than has usually been the
case. Equating grammar with form and the teaching of grammar with the teaching of explicit linguistic rules concerning form are unduly limiting,
representing what we have called myths (LarsenFreeman 1995), which only serve to perpetuate
the pendulum swing between language form and
language use. Grammar is about form and one
way to teach form is to give students rules; however, grammar is about much more than form,
and its teaching is ill served if students are simply
given rules.

251

Thus, in this chapter, we will entertain a

more robust view of grammar. Then, we will
briefly touch upon issues concerning its learning.
Finally, we will discuss its teaching.

A Three-Dimensional Grammar
Framework
Since our goal is to achieve a better fit between
grammar and communication, it is not helpful to
think of grammar as a discrete set of meaningless, decontextualized, static structures. Nor is it
helpful to think of grammar solely as prescriptive
rules about linguistic form, such as injunctions
against splitting infinitives or ending sentences
with prepositions. Grammatical structures not
only have (morphosyntactic) form, they are also
used to express meaning (semantics) in contextappropriate use (pragmatics). In order to guide
us in constructing an approach to teaching grammar that strives to meet this definition, it would
be helpful to have a frame of reference.
Our framework takes the form of a pie
chart. Its shape helps us to make salient that in
dealing with the complexity of grammar, three
dimensions must concern us: structure or form,
semantics or meaning, and the pragmatic conditions governing use.l Moreover, as they are
wedges of a single pie, we note further that the
dimensions are not hierarchically arranged as
many traditional characterizations of linguistic
strata depict.2 Finally, the arrows connecting one
wedge of the pie with another illustrate the interconnectedness of the three dimensions; thus a
change in any one wedge will have repercussions
for the other two.

FORM/
STRUCTURE

MEANING/
SEMANTICS

Morphosyntactic and
lexical patterns
Phonemic/graphemic
patterns

Lexical meaning
Grammatical
meaning

USE/
PRAGMATICS
Social context
Linguistic discourse context
Presuppositions about context

In the wedge of our pie having to do with

structure, we have those overt lexical3 and morphological forms that tell us how a particular
grammar structure is constructed and how it is
sequenced with other structures in a sentence or
text. With certain structures, it is also important
to note the phonemic/ graphemic patterns (see
the discussion of possessives and phrasal verbs
below for examples). In the semantic wedge, we
deal with what a grammar structure means. Note
that the meaning can be lexical (a dictionary definition for a preposition like down, for instance)
or it can be grammatical (e.g., the conditional
states both a condition and outcome or result).
It is very difficult to arrive at a definition of pragmatics distinct from semantics, and thus we are
sympathetic to Levinson's (1983) suggestion that
pragmatics deals with all aspects of meaning not
dealt with by semantic theory!
Since this definition is too broad for our
purposes here, however, we will limit pragmatics
to mean "the study of those relations between
language and context that are grammaticalized,
or encoded in the structure of a language"
(Levinson 1983, p. 9). We will leave the term context broad enough though, so that context can be
social (i.e., a context created by interlocutors,
their relationship to one another, the setting), or
it can be a linguistic discourse context (i.e., the
language that precedes or follows a particular
structure in the discourse or how a particular
genre or register of discourse affects the use of a
structure), or context can even mean the presuppositions one has about the context.
The influence of pragmatics may be ascertained by asking two questions:
1. When or why does a speaker/writer choose
a particular grammar structure over another
that could express the same meaning or
accomplish the same purpose? For example,
what factors in the social context might
explain a paradigmatic choice such as why a
speaker chooses a yes-no question rather
than an imperative to serve as a request for
information (e.g., Do you have the time?versus
Please tell me the time)?
2. When or why does a speaker/writer vary
the form of a particular linguistic structure?

252 Unit II E Language Skills/Grammar and Vocabulary

For instance, what linguistic discourse factors

would result in a syntagmatic choice such as
the indirect object being placed before the
direct object to create jenny gave Hank a
brand-new comb versus jenny gave a brand-new
comb to Hank?

POSSESSIVES

Despite the permeable boundaries between

the dimensions, we have found it useful to view
grammar from these three perspectives. We trust
that the utility of this approach will become
clearer as we proceed. A teacher of grammar
might begin by asking the questions posed in the
three wedges of our pie (for the sake of simplicity, labeled form, meaning, and use) for any given
grammar point.
Meaning of Possessive Besides possession, the
possessive or genitive form can indicate description (a debtor's prison), amount (a month's holiday),
relationship (Jack's wife), part/whole (my brother's
hand), and origin/agent (Shakespeare's tragedies).
Also, although all languages have a way of
signaling possession, they do not all regard the
same items as possessable. For example, Spanish
speakers refer to a body part using the definite
article instead of a possessive form. ESL/EFL
students will have to learn the semantic scope of
the possessive form in English.

Let us consider an example. A common

structure to be taught at a high-beginning level of
English proficiency is the 's possessive form. If we
analyze this possessive form as answers to our questions, we would fill in the wedges as below (analysis
based on Celce-Murcia and Larsen-Freeman 1999).
Form of Possessive This way of forming possessives in English requires inflecting regular
singular nouns and irregular plural nouns not
ending in s with 's or by adding an apostrophe
after the s' ending of regular plural nouns and
singular nouns ending in the sound /s/. This
form of the possessive has three allomorphs: /z/,
/s/, and /az/, which are phonetically conditioned:
/z/ is used when it occurs after voiced consonants
and vowels, /s/ following voiceless consonants,
and /az/ occurs after sibilants.

Use of Possessive Filling in this wedge requires

that we ask when the 's is used to express possession as opposed to other structures that can be
used to convey this same meaning. For example,
possession in English can be expressed in other
ways-with a possessive determiner (e.g., his, her,
and their) or with the periphrastic of the form (e.g.,
the legs of the table). Possessive determiners are presumably used when the referent of the possessor
is clear from the context. While ESL/EFL books
will often say that the of the possessive is used with
nonhuman head nouns and 's with human head
nouns, we are aware of certain conditions where
this rule does not apply. For example, native
speakers often prefer to use the 's even with inanimate head nouns if the head nouns are performing some action (e.g., the train's arrival was
delayed).4 Finally, students will have to learn to distinguish contexts in which a noun compound
(table leg) is more appropriate than either the 's
form or the of the form.

Teaching Grammar

253

Thus, by using our ternary scheme, we can

classify the facts that affect the form, meaning,
and use of the possessive structure. This is only a
first step. Teachers would not necessarily present
all these facts to students, recognizing that students can and do learn some of them on their
own. And certainly no teacher would choose to
present all these facts in a single lesson or on
one occasion. Nevertheless, distributing the features of the target grammatical structure among
the three wedges of the pie can give teachers an
understanding of the scope and multidimensionality of the structure. In turn, this understanding will guide teachers in deciding which
facts concerning the possessive will be taught
and when and how to do so.
Before continuing to explore these decisions, however, it might be worthwhile to apply
our approach to another grammar structure. Let
us analyze phrasal verbs this time. By considering
the three questions posed earlier, we can state the
following about phrasal verbs (analysis based
upon Celce-Murcia and Larsen-Freeman 1999):
PHRASAL VERBS

FORM
Verb + Particle (or)
Verb + Particle +
Preposition
Trans itive/Intransitive
Separable/Inseparable
Stress and juncture_
11 Patterns

MEANING
Literal
Figurative
Multiple Meanings

USE
Informal Discourse
Principle of Dominance

Form of Phrasal Verbs Phrasal verbs are two-part

verbs comprising a verb and a particle (e.g., to look
up). Sometimes, they can be constructed with
three parts in that a preposition can follow the
particle (e.g., to keep up with). As with all other
verbs, phrasal verbs are either transitive or intransitive. A distinctive feature of phrasal verbs is that
for many of them the particle can be separated
from its verb by an intervening object (e.g., Alicia

looked the word up in the dictionary). Phrasal verbs

also have distinctive stress and juncture patterns,
which distinguish them from verb plus preposition combinations:
Alicia looked up#the word.
Alicia walked#up the street.
Meaning of Phrasal Verbs There are literal
phrasal verbs, such as to hang up, where if one
knows the meaning of the verb or the particle or
both, it is not difficult to figure out the meaning of
the verb-particle combination. Unfortunately, for
the ESL/EFL student there are far more instances
of figurative phrasal verbs (e.g., to run into, meaning "meet by chance") where a knowledge of the
meaning of the verb and of the particle is of little
help in discerning the meaning of the phrasal
verb. Moreover, as with single-word verbs, phrasal
verbs can have more than one meaning (e.g., to
come across, meaning "to discover by chance" as in
I came across this old book in the library, or when used
intransitively, "to make an impression" as in
Richard came across well at the convention.
Use of Phrasal Verbs When is a phrasal verb
preferred to a single-word verb that conveys the
same meaning (e.g., put out a fire versus extinguish a fire)? For the most part, phrasal verbs
seem to be more common in informal spoken
discourse as opposed to more formal written discourse. When is one form of a phrasal verb preferred to another; i.e., when should the particle be
separated from its verb (e.g., put out afire versus
put afire out)? Erteschik-Shir's (1979) principle of
dominance seems to work well to define the circumstances favoring particle movement: If a noun
phrase (NP) object is dominant (i.e., a long, elaborate NP representing new information), it is
likely to occur after the particle; if the direct
object is short, old information (e.g., a pronoun),
it would naturally occur before the particle.
Identifying the Challenge Again, we would like
to underscore the fact that it would not be reasonable for the ESL/EFL teacher to present all of
this information to students at once. The framework does, however, help to organize the facts.
Furthermore, by doing this, teachers can more
easily identify where the learning challenge (s)

254 Unit 11 E Language Skills/Grammar and Vocabulary

will lie for their students. Identifying the challenging dimension ( s) is a key step which should
be taken prior to any pedagogical treatment.
All three dimensions will have to be mastered by the learner (although not necessarily
consciously). For phrasal verbs, it is the meaning
dimension which ESL/EFL students struggle
with most. It is often the fact that there is no systematic way of learning to associate the verb and
the particle. Adding to the students' woes, new
phrasal verbs are constantly being coined. By
recognizing where students will likely struggle,
an important clue is given the teacher as to
where to focus work on phrasal verbs. We will
amplify this point later. For now, however, it is
worth noting that although it is grammar structures which we are dealing with, it is not always
the form of the structures which creates the
most significant learning challenge.
"Grammaring" We should pause here to
acknowledge that as important as it is to develop
our understanding of the grammatical facts of the
language we are teaching, it is not these facts that
we wish our students to learn. We are not interested in filling our students' heads with grammatical paradigms and syntactic rules. If they knew all
the rules that had ever been written about
English but were not able to apply them, we
would not be doing our jobs as teachers. Instead,
what we do hope to do is to have students be able
to use grammatical structures accurately, meaningfully, and appropriately. In other words, grammar teaching is not so much knowledge
transmission as it is skill development. In fact, it
is better to think of teaching "grammaring"
(Larsen-Freeman 1997; 2001), rather than
"grammar." By thinking of grammar as a skill to
be mastered, rather than a set of rules to be
memorized, we will be helping ESL/EFL students go a long way toward the goal of being able
to accurately convey meaning in the manner they
deem appropriate.

The Learning Process

However important and necessary it is for teachers to have a comprehensive knowledge of their
subject matter, it is equally important for them

to understand their students ' learning process.

This understanding can be partly informed by
insights from second language acquisition
(SLA) research concerning how students naturally develop their ability to interpret and produce grammatical utterances . Three insights are
germane to our topic:
1. Learners do not learn structures one at a
time. It is not a matter of accumulating structural entities (Rutherford 1987).5 For example, it is not the case that learners master the
definite article, and when that is mastered,
move on to the simple past tense. From their
first encounter with the definite article,
learners might master one of its pragmatic
functions-e.g., to signal the uniqueness of
the following noun. But even if they are able
to do this appropriately, it is not likely that
they will always produce the definite article
when needed because learners typically take
a long time before they are able to do this
consistently. Thus, learning is a gradual
process involving the mapping of form, meaning, and use; structures do not spring forth in
learners' interlanguage fully developed and
error-free.
2. Even when learners appear to have mastered
a particular structure , it is not uncommon to
find backsliding occurring with the introduction of new forms to the learners' interlanguage. For example, the learner who has
finally mastered the third person singular
marker on present-tense verbs is likely to overgeneralize the rule and apply it to newly
emerging modal verbs, thus producing errors
such as She cans speak Spanish. Teachers should
not despair, therefore, at regressive behavior
on the part of their students. Well-formedness
is usually restored once the new additions
have been incorporated and the system selforganizes or restructures.
3. Second language learners rely on the knowledge and the experience they have. If they are
beginners, they will rely on their LI as a
source of hypotheses about how the L2 works;
when they are more advanced, they will rely
increasingly on the L2. In understanding this,
the teacher realizes that there is no need to

Teaching Grammar

255

teach everything about a structure to a

group of students; rather, the teacher can
build upon what the students already know.
It also follows that the challenging dimension for a given grammatical structure will
shift from class to class depending on the
students' L1 backgrounds and level of L2
proficiency. Successful teaching involves
identifying the relevant challenge for a particular group of students.
To these three observations, we will add a
fourth one that is not to our knowledge treated in
the SLA research literature, but rather one based
upon our observations and supported by learning
theorists (e.g., Gagne and Medsker 1996).
4. Different learning processes are responsible
for different aspects of language. Indeed,
given that language is as complicated as it is,
one would not expect the learning process to
be any simpler. It is clearly an oversimplification to treat all grammar learning as resulting
from habit formation or from rule formation.
Being aware that different learning processes
contribute to SLA suggests a need for the
teaching process to respect the differences.
How the nature of the language challenge
and the learning process affect teaching decisions is the issue to which we turn next.

The Teaching Process

Consistent with the way we are conceiving grammar in this chapter, teaching grammar means
enabling language students to use linguistic forms
accurately, meaningfully, and appropriately. In
this section we discuss various teaching strategies
that can be employed to meet this goal.
In keeping with language form approaches,
traditional grammar teaching has employed a
structural syllabus and lessons composed of three
phases: presentation, practice, and production
(or communication), often referred to as "the
PPP" approach. As we saw earlier, underlying this
approach is the assumption that one systematically builds towards communication. However,
as mentioned in the introduction to this chapter, these days, most teachers embrace a more

communicatively oriented approach , starting

with a communicative activity such as taskor content-based material ( see chapters by
Savignon and Snow in this volume). The grammar that is taught is not scheduled in advance as
it is with a structural syllabus / PPP approach, but
rather supports students in their completion of
the communicative task or their making sense of
a particular content area . In addition , or alternately, teachers respond to grammar errors that
students commit when engaged in communication. As such , it reverses the normal sequence
(Skehan 1998b ), putting communication first,
rather than selecting and presenting a grammar
structure in advance of its use in context.
Even if the grammar to be worked on is
derivative rather than scheduled in advance, a
teacher must still decide how to address it. A
variety of options have been suggested (see
Doughty and Williams 1998; R. Ellis 1998),
although the research findings underpinning
them are somewhat sparse and sometimes contradictory ( see Mitchell 2000 for a recent review).
One option is simply to bring to students ' attention , or to promote their noticing of, some feature
of a grammatical structure . For example , if a student makes an error and the teacher decides to
respond to it, then the teacher might recast or
reformulate what the student has said or written
incorrectly in a more accurate , meaningful, or
appropriate manner. For instance , if it is an error
of form, the teacher would recast the student's
production accurately.
STUDENT:
TEACHER:

This is Juan notebook.

Oh. That is Juan 's notebook.
(perceiving the error to be the
form of the possessive)

If meaning is the problem, the teacher

would recast what the student has said in a
meaningful way.
STUDENT: I need to look at the word in the
dictionary.
TEACHER: You need to look up the word
in the dictionary.
(perceiving the phrasal verb
look up to be a better form for
what the student means to say)

256 Unit II E Language Skills/Grammar and Vocabulary

And if use is the problem, the teacher

would recast what the student has said in a more
appropriate manner:
STUDENT: I arise at six in the morning.

TEACHER: OK. You get up at six in the

morning.
(perceiving that a phrasal verb would
be more appropriate to convey the
student's intended meaning)
A more proactive way to promote students'
noticing a particular grammatical structure is to
highlight it in a text in some fashion. Enhancing
the input (Sharwood Smith 1993) might be an
especially effective way to focus students' attention on grammar structures that operate at the
discourse level of language, such as articles or
verb tenses. By boldfacing all the normally
insalient articles in a given passage, for instance,
the students' attention could be drawn to them.
Even simply choosing texts in which a particular
structure or structural contrast is especially frequent would enhance its saliency and thus might
promote noticing, a practice sometimes called
input flooding.

Still another option is to use a consciousnessraising task, in which it is the students' job to
induce a grammatical generalization from the
data they have been given. For example, Fotos
and Ellis (1991) ask students to work out the
rule for indirect object alternation in English
(e.g., They gave a gold watch to him./They gave him
a gold watch.) by giving the students example sentences where indirect object alternation can and
cannot be successfully applied. Indirect object
alternation is difficult in English and therefore is
an ideal candidate for this sort of explicit rule
articulation. Indeed, Carroll and Swain (1993)
suggest that when the rules are not that clear-cut,
detailed instruction with explicit metalinguistic
feedback may be the most helpful response to
student errors.
Another option for promoting students'
awareness is to use the garden path strategy
(Tomasello and Herron 1988; 1989). As applied
to grammar teaching, this means giving students
information about structure without giving
them the full picture, thus making it seem easier
than it is, or in other words, "leading them down
Teaching

the garden path." If ESL/EFL students were told

that the English past tense is formed with -ed, for
example, this would be leading students down
the garden path as there are many irregular
verbs in English where this rule will not work to
produce the past tense. The reason for giving
students only a partial explanation is that they
are more likely to learn the exceptions to the
rule if they are corrected at the moment the
overgeneralization error is made than if they are
given a long list of "exceptions to the rule" to
memorize in advance.
Another technique for directing students'
attention to form is called input processing (Van
Patten 1996). Rather than working on rule
learning and rule application, input processing
activities push learners to attend to properties of
language during activities where the structure is
being used meaningfully. For instance, if students are asked to carry out commands that
teachers issue, they are working on matching the
imperative form to its use in a meaningful way.
Of course, sometimes a communicative task
itself requires that students attend to relevant
features of the target language (Loschky and
Bley-Vroman 1993), such as when using a particular grammatical structure is essential to completing the task. An example of this is when
students have to use particular prepositions to
accurately give each other directions using a
map. The added value of using a communicative
task to promote noticing is that students are
encouraged to use the target structures, thereby
generating "output" that attracts feedback from
a teacher or another student.
Speaking of output, it might be surprising to
experienced teachers to read descriptions of all
these teaching options with very little mention of
student production. But, of course, students' production plays a very important role in learning
grammar. It is not enough to have awarenesses
raised if students can't produce the language.
Output production is, therefore, extremely important. For one thing, it pushes students to move
beyond semantic processing to syntactic processing (Swain 1985). Then, too, when students
attempt to produce structures, they get to test their
hypotheses on how the structure is formed or what
it means or when it is used. Following these

Grammar

257

attempts, as we have seen, they can receive feedback on their hypotheses and modify them as
necessary.

Indeed, Donato (1994) has shown how students' participation in collaborative dialogue,
through which learners can provide support for
each other, has spurred development of learners'
interlanguage. Other research (Swain and Lapkin
1998) corroborates the value of an interactive dialogue as both a cognitive tool and a means of
communication which can promote grammatical
development.
Beyond these reasons for giving students an
opportunity to produce the target grammatical
structures, we have already presented the idea
that grammar teaching can better be thought of
as developing "grammaring," i.e., helping students be able to use grammar skillfully, a goal
that requires significant practice. To this point,
Gatbonton and Segalowitz (1988) have argued
that practice of grammatical patterns can lead to
automatization of certain aspects of performance, which, in turn, frees up students' attentional resources to be allocated elsewhere.
It used to be that the practice phase of a lesson was devoted almost exclusively to grammar
drills and exercises. Ever since the ineffectiveness of using drills which do not engage students' attention was acknowledged, there has
been little by way of guidance offered on how to
give students meaningful practice. What follows,
therefore, is an attempt to fill this void. Practice
activities will be addressed in terms of which
dimension of language they relate to.
Form
From what we know of skill acquisition theory
(e.g., Anderson and Fincham 1994), fluency or
proceduralization of declarative knowledge
(e.g., knowledge of a grammar rule or pattern)
requires practice in which students use the target
language point meaningfully while keeping the
declarative knowledge in working memory
(DeKeyser 1998).
It is important to emphasize meaningful practice of form for several reasons. First of all, meaningless mechanical drills, such as repetition drills,
commonly associated with behaviorist approaches
to learning, do not engage the learner in the
258

target behavior of conveying meaning through

language. Furthermore, because students are not
engaged in target behavior, the inert knowledge
problem (Whitehead 1929) is likely to materialize.
Inert knowledge is knowledge that can be recalled
when students are specifically asked to do so but is
not available for spontaneous use, in, say, problem
solving, even when the knowledge is relevant to
the problem at hand. Knowledge remains inert
when it is not available for transfer from the classroom context to the outside world. We know that
when the psychological conditions of learning and
application are matched what has been learned is
more likely to be transferred (e.g., Blaxton 1989).
Thus, rules and forms learned in isolated meaningless drills may be harder to retrieve in the context of communicative interaction (Segalowitz and
Gatbonton 1994). Finally, student motivation
is likely to be enhanced if students are able to
interact in a way that is meaningful to them. Then,
too, they are likely to be more attentive if they are
saying something meaningful.
Identifying the type of learning involved
helps us to think about the desirable characteristics of any practice activity. For instance, for declarative knowledge to be proceduralized a great deal
of meaningful practice would be required. Further,
students would have to receive feedback on the
accuracy with which they produced the target
form. They would have to be restricted to using
just the particular target form; in other words,
structural diversity would not be permitted.6
Finally, for proceduralization to occur, it would
seem important to concentrate on only one or two
forms at a time, although, of course, the target
form could be introduced in contrast to forms
that the student already controls.
Let us take an example and see how these
characteristics are applied. If our students show
us that they are struggling with the inversion of
the subject and operator in yes-no questions, it
would be clear that their immediate learning challenge is linguistic form. We will need to design or
select an activity that encourages meaningful practice of the pattern, not verbatim repetition. We
want the students to concentrate on producing
only yes-no questions. A game like Twenty
Questions would appear to meet the criteria.
Students get to ask 20 yes-no questions about an

Unit II E Language Skills/Grammar and Vocabulary

object or person in an attempt to guess the identity; hence, they receive abundant practice in
forming the questions, and the questions they
produce are meaningful. The teacher would
work with each student to enable him or her to
produce the pattern accurately, perhaps providing an explicit rule, perhaps not.
An example of a game that would work on
the form of the English possessive comes from
Kealey and Inness (1997). Students are given a
family portrait in which the child's face is missing.
They are also given clues as to what the child looks
like, e.g., the child has the mother's eyebrows or
the father's chin. A person from each small
group into which the students are divided comes
to the front of the room, takes a clue, memorizes
it, and brings it back to his or her group so that
the feature in the clue can be drawn. This continues until the child, a composite of his mother and
father, is fully drawn.
In sum, certain games are good devices for
practicing grammar points where the challenge
resides in the formal dimension. While not an
activity in and of itself, another useful device for
working on the formal dimension is the use of
cuisenaire rods. The rods are ideal for focusing
student attention on some syntactic property
under scrutiny. One example that comes to mind
is an adaptation of Stevick's (1980) Islamabad
technique. Practicing the form of OS7 relative
clauses, students might be asked to use the rods
to construct a view of some spot in their hometown. The students would be encouraged to use
OS relative clauses where appropriate (e.g., There
is a fountain that is located in the center of my town;
Around the fountain there are many people who sell
fruits, vegetables, and flowers, etc.).

One final example of a type of useful activity

for working on the formal dimension is a problemsolving activity. The problem to be solved could be
most anything, but if we are dealing with the formal dimension, then we would want it to conform
to the characteristics described above. An example
might be an information-gap activity where the
students are given a class information sheet with
certain items missing (see bottom of this page).
Students could circulate asking one another
Wh-questions (e.g., What is Beatriz's major? How old
is Werner?) in order to complete the chart.
Another example might be a sentence-unscrambling task. This is a useful problem-solving activity
when the challenge is getting students to produce
correct word order, such as when the objective is
to have students use auxiliary verbs in the proper
sequence.
It is important to take note that there is
nothing inherent in the three examples we have
provided (games, use of rods, problem-solving
activities) which make them useful for addressing the formal dimension; i.e., we could easily
use rods to work on some aspects of the meaning
or use dimensions. What is significant to remember is that the activity should be structured in
such a way that it is compatible with the characteristics presented earlier.
Meaning
If the teacher has decided that the challenge of a
particular structure lies in the semantic dimension, then a different sort of practice activity
should be planned. It would seem that meaning
would call for some sort of associative learning
(N. Ellis 1998), where students have opportunities

Name

Age

Country

Language

Major

Beatriz

18

Bolivia

Spanish

Dentistry

Mohammed

19

Algeria
France

Jean Claude
18
Werner

17

Hobby

Accounting

Painting

French
Education

Brazil
Swiss German

Going
to the movies

Hiking

Business

Teaching Grammar

259

to associate the form and the meaning of the

particular target structure. It has been our experience that repetition is not needed to the same
extent as it is when teaching some aspect of the
formal dimension. Sometimes a single pairing
of form and meaning suffices. Due to memory
constraints, it seems prudent to restrict the number of new items being practiced at any one time
to between two and six (Asher 1996). The students would receive feedback on their ability
to demonstrate that they had acquired the formmeaning bond.
Celce-Murcia and Hilles (1988) mention
that when dealing with the semantic dimension,
realia and pictures are very useful. Thus, for
example, if the teacher has decided to work on
the semantics of comparative forms in English to
support some communicative task or content, he
or she might show students pairs of pictures and
work with them to make comparisons using the
form that reflects the relation depicted (e.g., as
as, more than, less than).
Actions, too, can make meaning salient.
The initial challenge for ESL/EFL students
grappling with prepositions is to associate the
"core" meaning with each. Thus, prior to having
students work on direction-giving tasks using
maps (Walk to the corner. Turn right at the corner. The
cinema is near the corner, next to the bank.), a good
strategy might be to work with students on having
them make an association between a preposition
and its meaning in locating objects in space. One
way of doing this is to conduct a Total Physical
Response sequence where students act out a
series of commands along with the teacher,
involving the placement of objects in various parts
of the room; e.g., Put the book next to the desk, Put
the pen on the book, Walk to the door, Stand near the
door, etc. Once students appear to have made the
connection between form and meaning, the
teacher can assess their ability to discriminate one
form-meaning bond from another by having
them carry out commands on their own and by
issuing novel commands-e.g., Put the pen on the
desk-and assessing their ability to comply.
We said earlier that a persistent challenge
for students' learning phrasal verbs was the fact
that the meaning is often not detectable from
combining the meaning of the verb with the

meaning of the particle. Sometimes teachers have

had their students play Concentration, a version
of the game in which the students have to associate a phrasal verb written on one card with its definition written on another card. Another example
of an activity that would address this semantic challenge is an operation (Nelson and Winters 1993).
In an operation, a series of separate actions are
performed to accomplish some task. The teacher
might issue commands, or mime the actions with
the students as she or he describes them.
I want to call up my friend. First, I look up
the phone number Then I write it down. I
pick up the receiver and punch in the number. The number is busy. I hang up and
decide to call back later.

By practicing this operation several times,

the students can learn to associate the form and
meaning of certain phrasal verbs (call up, look up,
pick up, etc.). If students are given an operation
with which to associate phrasal verbs, recall at a
later time will likely be enhanced. To determine
if students can distinguish among the various
phrasal verbs, students might be given phrasal
verbs out of sequence and asked to mime the
appropriate action. Feedback on their ability to
match form and meaning can be given.
Use
When use is the challenge, it is because students
have shown that they are having a hard time
selecting the right structure or form for a particular context. Working on use will involve students learning that there are options to be
exercised and that they must select from among
them the one which best suits a given context.
Thus, relevant practice activities will provide
students with an opportunity to choose from two
or more forms the one most suitable for the context and how they wish to position themselves
(e.g., in a cooperative way, a polite way, an assertive
way, etc.). Students would receive feedback on the
appropriateness of their choice. In some cases,
their choice might involve selecting between two
options (e.g., when to use the passive versus the
active voice). Other times, their choice would be
from among an array of options (e.g., which

260 Unit II E Language Skills/Grammar and Vocabulary

modal verb to use when giving advice to a boss);

hence, the number of forms being worked on at
one time would be at least two, but could involve
many more.

Role plays work well when dealing with use

because the teacher can systematically manipulate social variables (e.g., increase or decrease the
social distance between interlocutors) to have students practice how changes in the social variables
affect the choice of form.
For example, if students have shown that
they do not know how to use modals to give
advice, they might be asked to role-play having a
"dilemma." In this role play, one person has a
problem; (e.g., the keys to the car have been
lost. The car is locked and the person wants to
get in.) Students are asked to use modal verbs to
give advice to the person with the problem; e.g.,
You might try breaking the window, You could try calling the police. The teacher could next alter a
salient feature of the context, thus creating a new
social context in which a different modal verb
would be more appropriate. For example, the
teacher might ask, "What if it were a young child
that had this dilemma?" A more appropriate
form and content for the advice, then, might be
You had better wait for your mother to come!
On another occasion, students might be
asked to play the role of an advice columnist.
They are to write a column and give advice to a
classmate who is having a particular problem.
Having students work with the same structure in
writing and in speaking activities can highlight
differences between written and oral grammars
(Carter and McCarthy 1995).
Role plays are useful for highlighting other
structural choices as well. Often we find that it is
neither the form nor the meaning of the English
tenses that presents the greatest long-term challenge to ESL/EFL students; rather it is when/why
to use one tense and not the other. In other
words, it is the pragmatic usage of the tenses that
is the major obstacle to their mastery. Giving students practice with situations in which a contrast
between two tenses is likely to arise may sensitize
students to the usage differences. For instance, a
notorious problem for ESL/EFL students is to
know when to use the present perfect versus
when to use the past tense. A situation where a
Teaching

contrast between them would occur might be a job

interview. In such a context, the perfect of experience is likely to be invoked (e.g., Have you ever done
any computer programming?). An elaboration to an
affirmative answer is likely to contain the past tense
(e.g., Yes, I have. I once worked on ... or simply, Yes.
When I worked at ... ). Students can take turns roleplaying the interviewer and interviewee.
As was mentioned earlier, it is not only the
social context that will be involved in the choice of
which forms to use, but also it is often the linguistic discourse context that will make a difference.
Thus, it is very important to consider teaching discourse grammar (Celce-Murcia 1991a; Hughes
and McCarthy 1998). Such is the case with the passive voice. Its use is not particularly sensitive to
social factors; i.e., whether one is using the active
or passive voice does not necessarily depend upon
with whom one is conversing. What usually does
cause students considerable difficulty with the passive voice, however, is determining when to use it.
The fact that the agent of an action is defocused
motivates the use of the passive. Furthermore, if
the agent has already been established in the linguistic discourse, it would likely not even be mentioned in subsequent discourse. Thus, most passive
sentences are agentless.
Challenges of this nature call for textgeneration or text-manipulation-type exercises.
As the passive is used more often in written than
in spoken English, teachers might give their students a text-completion exercise in which the
first few lines of the text are provided. For example, from the first few lines in the following text,
it should be clear to the students that the theme
of discourse is on the "issues," not the agents (i.e.,
participants), at the town meeting.
Town meetings were held throughout
New England yesterday. Many issues
were discussed, although the big one
for most citizens was the issue of
growth. Many changes have been
made recently. For example, .. .
Students then are asked to complete the text
using the appropriate voice. As not all the sentences should be in the passive voice, students will
be making choices, in keeping with a characteristic
of practice activities designed to work on the use

Grammar

261

dimension. The teacher will give feedback

to the students on the appropriateness of their
choices.
Before leaving our discussion of the passive
voice, it would be useful to illustrate why we feel
that identifying the challenging dimension is a
worthwhile step to take before teaching any
grammar structure. When we are clear where
the challenge lies, the challenge can shape our
lessons. For instance, as we stated earlier, it has
been our experience that the greatest long-term
challenge for students working on the passive
voice is for them to figure out when to use the
passive. Keeping this in mind will help us avoid a
common practice of ESL/EFL teachers, which is
to introduce the passive as a transformed version
of the active (e.g., "Switch the subject with the
direct object . . ."). Presenting the passive in this
way is misleading because it gives the impression
that the passive is simply a variant of the active.
Moreover, it suggests that most passive sentences
contain agents. What we know in fact to be the
case is that one voice is not a variant of the other,
but rather the two are in complementary distribution, with their foci completely different. We
also know that relatively few passive sentences
contain explicit agents. Thus, from the first, the
passive should be taught as a distinct structure
which occurs in a different context from the
active. (See Celce-Murcia and Larsen-Freeman
1999, for several examples of how to do this.)
It should be noted that the pie chart, the
observations about learning, and the characteristics of practice activities enumerated here may not
significantly alter the way grammar is taught today.
Indeed, many of the activities recommended here
are currently being used. What these tools do offer,
how-ever, is a principled means for dealing with
grammar. They should help teachers to make clear
decisions they teach grammar. They should help
teachers to design effective activities or to choose
from among those in a textbook without assuming
that just because a textbook activity deals with the
target structure, it necessarily addresses the particular learning challenge that their students are
experiencing.
This brings us to the close of our discussion
on how to design practice activities for grammar
points.

Providing Feedback
Providing learners with feedback, negative evidence which they can use to correct their misapprehensions about some aspect of the target
language, is an essential function of language
teaching. Even such indirect feedback as asking a
learner for clarification of something he or she
has said may be helpful (Schachter 1986). It has
always been a controversial function, however
(Larsen-Freeman 1991). There are, for instance,
those who would proscribe it, believing that a
teacher's intervention will inhibit students from
freely expressing themselves or that there is little
evidence demonstrating that learners make use of
the feedback they have been given-there is little
immediate "uptake" of the correct form. While
there are clearly times that such intervention can
be intrusive and therefore unwarranted (e.g., in
the middle of a small-group communicative activity), at other times focused feedback is highly
desirable. Further, immediate uptake cannot be
the sole criterion of its usefulness. Negative evidence gives students the feedback they need to
reject or modify their hypotheses about how the
target language is formed or functions. Students
understand this, which explains why they often
deliberately seek feedback.
The same pie chart that we used when identifying the learning challenge and creating
practice activities can also be a useful aid in diagnosing errors. When an error is committed by a
student, a teacher can mentally hold it up to the
pie chart to determine if it is an error in form,
meaning, or use. Of course, sometimes the cause
of an error is ambiguous. Still, the pie chart does
provide a frame of reference, and if the diagnosis
is accurate, the remedy may be more effective.
More than once we have observed a teacher give
an explanation of linguistic form to a student,
when consulting the pie chart would have
suggested that the student's confusion lay with
the area of use instead.
As for how the feedback is to be provided, we
have already mentioned several useful optionsrecasting, for instance. Getting students to selfcorrect is another (see Lyster and Ranta 1997).
Giving students an explicit rule is a third. Some
teachers like to collect their students' errors,

262 Unit II E Language Skills/Grammar and Vocabulary

identify the prototypical ones, and then deal

with them collectively in class in an anonymous
fashion. Which of these options is exercised will
depend on the teacher's style, the proficiency of
the students, the nature of the error, and in which
part of the lesson the error has been committed.
None of these have to be used exclusively,
of course. For instance, Aljaafreh and Lantolf
(1994) offer a graduated 12-point scale ranging
from implicit to explicit strategies, beginning
with student identification of errors in their own
writing, moving to where the teacher isolates the
error area and inquires if there is anything
wrong in a particular sentence, to where the
teacher provides examples of the correct pattern
when other forms of help fail to lead to a selfcorrection on the part of the student.

RELATED PEDAGOGICAL
ISSUES
Sequencing
Earlier we noted that grammar structures are
not acquired one at a time through a process of
"agglutination" (Rutherford 1987). Rather, different aspects of form, meaning, and use of a
given structure may be acquired at different
stages of L2 development. This observation confirms the need for recycling-i.e., working on
one dimension of a form and then returning to
the form from time to time as the need arises.
To some extent this will occur naturally, as the
same structures are likely to be encountered in
different communicative tasks and content areas.
However, it is also the case that not all linguistic
structures that students need to learn will be available in the language that occurs in the classroom.
Therefore, it will be necessary for the
teacher to "fill in the gaps," i.e., to introduce
structures that don't naturally arise in classroom
discourse (Spada and Lightbown 1993). For this
reason, teachers might think in terms of a grammar checklist, rather than a grammatical sequence.
By this, I mean that it would be a teacher's responsibility to see that students learn certain grammatical items by the end of a given course or

period of time, but not by following a prescribed

sequence. Many structures would arise naturally
in the course of working on tasks and content and
would be dealt with then. Other structures might
be introduced as the teacher determined that the
students were ready to learn them. Rather than
adhering to a linear progression, the choice of
sequence would be left up to the teacher and
would depend on the teacher's assessment of the
students' developmental readiness to learn.
Many teachers, of course, have little control
over the content or sequence of what they work
on. They must adhere to prescribed syllabi or
textbooks, although even in such a situation, it
may be possible for teachers not to follow a
sequence rigidly. But for those teachers who have
more flexibility, research on acquisition orders is
germane. Some SLA research has shown that
learners progress through a series of predictable
stages in their acquisition of particular linguistic
forms. One explanation for the order rests on
the complexity of the speech-processing strategies required. Thus, all structures processable by
a particular strategy or cluster of strategies
should be acquired at roughly the same developmental stage. This approach has been shown to
account for certain acquisition orders in ESL
(Pienemann and Johnston 1987).
Despite these findings and their potential
implications for grammatical structure sequencing, there has been no definitive acquisition
order established, and thus teachers are still left
to their own resources for judgments on how to
proceed. We should also note that even if an
acquisition order were to be fully specified for
English, there might be justification for preempting the acquisition order when students' communicative needs were not being met and when,
therefore, certain structures would need to be
taught, at least formulaically. Furthermore,
Lightbown (1998) has suggested that even if students are asked to work on structures before they
are ready to acquire them, such effort may not be
in vain because such instruction might prime subsequent noticing on the part of the students,
thereby accelerating acquisition when they are
indeed ready.

Teaching Grammar

263

Inductive Versus Deductive

Presentation
An additional choice teachers face is whether to
work inductively or deductively. An inductive
activity is one in which students infer the rule or
generalization from a set of examples. In a deductive activity, on the other hand, the students are
given the rule and they apply it to examples. For
instance, when practicing an inductive approach
to the mass/count noun distinction in English,
students could be presented with a language sample, such as a grocery advertising circular. They
then would be encouraged to make their own
observations about the form of mass and count
nouns. The teacher might listen to their observations and then might summarize by generalizing
about the two categories of nouns. If practicing a
deductive approach, the teacher would present
the generalization and then ask students to apply
it to the language sample.
As we see, if a teacher has chosen an inductive approach in a given lesson, a further option
exists-whether or not to give or have students
articulate an explicit rule. Earlier, we stated that
equating the teaching of grammar with the provision of explicit rules was an unduly limited
view of what it means to teach grammar. We said
this because what we are trying to bring about in
the learner is linguistic behavior that conforms
to the rules, not knowledge of the rules themselves. Having said this, we see no reason to
avoid giving explicit rules as a means to this end,
except perhaps if one is working with young children. Usually students request rules and report
that they find them helpful. Moreover, stating a
rule explicitly can often bring about linguistic
insights in a more efficacious manner, as long as
the rule is not oversimplified or so metalinguistically obtuse that students must struggle harder
to understand the rule than to apply it implicitly
(Robinson 1996).
Returning now to the inductive versus
deductive question, we again find that the choice
is not one resolvable with an either/or approach.
There are many times when an inductive
approach such as using a consciousness-raising
task is desirable because by using such an

approach one is nurturing within the students a

way of thinking, through which they can arrive at
their own generalizations. In addition, an inductive approach allows teachers to assess what the
students already know about a particular structure and to make any necessary adjustments in
their lesson plan. Clearly, a teacher's anticipation
of where the challenge lies is not always borne out
when he or she assesses students' actual behavior.
Other times, when students have a particular
cognitive style that is not well suited for language
analysis or when a particular linguistic rule is
rather convoluted, it may make more sense to
present a grammar structure deductively.
Indeed, Corder's sensible observations
offer comfort:
What little we know about ... second
language learning ... suggests that a
combination of induction and deduction produces the best result.... The
old controversy about whether one
should provide the rule first and then
the examples, or vice versa, is now
seen to be merely a matter of tactics to
which no categorical answer can be
given (Corder 1973 in Rutherford and
Sharwood Smith 1988, p. 133).
Patterns and Reasons , Not Rules
Before concluding, we should make two final
observations about grammar teaching. With the
increased access to large corpora of language
data that computers afford, it has become clear
that grammatical structures and lexical items
occur in a large number of regularly occurring
patterns (Sinclair and Fox 1990; Biber, Conrad,
and Reppen 1998). Not all lexical items can be
freely substituted into a particular pattern. Once
one lexical item is selected, the likelihood of a
particular item or phrase following is increased.
For example, if the verb insist is chosen, either on
or that is very likely to follow. An implication of
corpus-based research is that teachers of grammar should pay more attention to conventionalized lexicogrammatical units, and not simply
focus on teaching grammatical rules (Pawley
and Syder 1983; Nattinger and DeCarrico 1992;

264 Unit II E Language Skills/Grammar and Vocabulary

Lewis 1997). Indeed, connectionist modeling

has demonstrated that morphology (Ellis and
Schmidt 1997) and syntax (MacWhinney 1997)
acquisition may be accounted for by simple associative learning principles (N. Ellis 1998), rather
than as a product of rule application.
Another challenge to equating the teaching of rules with the teaching of grammar comes
from Larsen-Freeman's (2000a) suggestion that
teachers concentrate on teaching "reasons, not
rules." Larsen-Freeman points out that although
rules don't allow for change, language is changing all the time. A consequence is that most rules
have "exceptions." Furthermore, many rules
appear arbitrary because they are form based,
ignoring the meaning and use dimensions. For
instance, rather than telling students they must
use an indefinite noun phrase after the verb in a
sentence beginning with existential there,
There is a snowstorm coming.
help them understand the reason: there introduces new information in the noun following
the verb, and in English, new information is
marked with indefinite determiners. This reason
is broad based and explains a number of English
word-order phenomena. While rules provide
some security for learners, reasons give them a
deeper understanding of the logic of English
and help them make it their own. Besides, reasons are meaning based and use based and are
in keeping with the more robust view of grammar we have been promoting in this chapter.

sions of grammar and how to teach them, teachers will continue to develop their professional
knowledge base, which will, in turn, benefit their
students as they strive to enhance their grammatical proficiency.

DISCUSSION QUESTIONS
1. Think of all the language teaching approaches with which you are familiar. Can
you categorize them according to whether
they favor language form or language use?

2. In explaining the pragmatics of phrasal

verbs, the principle of dominance was
invoked. Explain why the principle of dominance falls in the pragmatic dimension.
3. The effect of the native language on second
language learning has traditionally been
seen to be one of interference. How does
observation 3 on the learning process
(pp. 255-256) differ in its perception of
L1 influence?
4. Why was it stressed that the repetition in a
practice activity for working on form should
be meaningful?
5. Why is it important to identify the challenge
in a particular grammar structure for a particular group of students, even if the aspect
of structure you are planning to teach lies in
a different wedge of the pie from where the
challenge lies?

SUGGESTED ACTIVITIES
PROFESSIONAL DEVELOPMENT
Finally, the form, meaning, and framework can
be used by teachers to assess where there are
gaps in their own knowledge of English grammar. When they can't fill in all the wedges in the
pie chart for a given structure, they can consult
reference grammars. Of course, there are many
gaps in what is known about the three dimensions. In particular, there is much to learn about
the pragmatic conditions governing the use of
particular structures. For this reason, the pie
chart can also be used to generate items for a
research agenda. By exploring the three dimen-

1. Think of a language teaching approach

which tends to favor language use over language form. How could the approach incorporate more language form? Now think of
an approach that favors language form
over language use. How could a focus on
language use be integrated?
2. Analyze restrictive relative clauses in terms
of the three dimensions of the pie chart.
What has been the most challenging dimension for the students with whom you have
worked?
3. Design practice activities for dealing with the
pragmatics of the following:

Teaching Grammar

265

a. falling versus rising intonation in tag

questions
b. indirect object alternation
c. presence or absence of existential there

OQ!^ FURTHER READING

Bygate, M., A. Tonkyn, and E. Williams, eds. 1994.
Grammar and the Language Teacher. Hemel
Hempstead. UK: Prentice Hall International.
Offers ways that grammar in language teaching
can be reaffirmed and maintained in order to
avoid the pendulum swing.
Celce-Murcia, M., and S. Hilles. 1988. Techniques and
Resources in Teaching Grammar. New York: Oxford
University Press.
Discusses issues germane to teaching grammar
and provides abundant examples of techniques
and materials applied to teaching English
structures.
Celce-Murcia, M., and D. Larsen-Freeman. 1999. The
Grammar Book: An ESL/EFL Teacher's Course. 2d
ed. Boston, MA: Heinle & Heinle.
Seeks to guide teachers to an understanding of
the grammar of those structures they will have
to teach (their form, meaning, and use in context) and offers relevant teaching suggestions
for those same structures.
Doughty, C., and J. Williams, eds. 1998. Focus on Form
in Classroom Second Language Acquisition.
Cambridge: Cambridge University Press.
Provides an overview of second language acquisition research that has investigated "focus on
form."
Larsen-Freeman, D. 2001. Teaching Language: From
Grammar to Grammaring. Boston, MA: Heinle &
Heinle.
Argues for a reconceptualization of grammar
and the way it is taught, featuring grammar as a
complex, nonlinear, dynamic system.
Rutherford, W. 1987. Second Language Grammar:
Learning and Teaching. London: Longman.
Treats grammar in an interesting and provocative way that challenges the view that learning
grammar is an "accumulation of entities."
Ur, P. 1988. Grammar Practice Activities: A Practical Guide
for Teachers. Cambridge: Cambridge University
Press.
Discusses pedagogical issues followed by a number of grammar teaching activities grouped
according to the grammar structure for which
they work best.

ENDNOTES
I

Some time after I had begun view grammar in this

way, the work of Charles Morris (1939) was brought
to my attention. Although he uses the terms in a
somewhat different manner, Morris applies the ternary scheme of syntactics, semantics, and pragmatics in portraying the field of semiotics or the study
of signs. The ternary scheme we are adopting here
may also sound reminiscent of Kenneth Pike's
"particle, wave and field" (1959). Although there is
some overlap, there is no isomorphism between the
models.
2 For example, the model of language that descriptive linguists prefer is one in which various areas of
language are depicted as strata in a linguistic hierarchy, beginning with the sounds of language as
the lowest level from which all else is composed
and following in turn with morphemes, lexicon,
syntax, and discourse.
3 We include lexis here, acknowledging that grammar and lexis are just two poles on a continuum
and that there are many patterned multiword
phrases that are basic intermediate units between
lexis and grammar. Following Halliday (1994),
then, it is probably more accurate to think in terms
of "lexicogrammar."
4 For more exceptions to this rule, consult CelceMurcia and Larsen-Freeman (1999, pp. 314-316).
5 For this reason, Rutherford has suggested that an
optimal approach to dealing with the nonlinearity
of grammatical acquisition might be one where
teachers help students achieve an understanding
of general principles of grammar, e.g., how to
modify basic word order, rather than concentrating on teaching structure-specific rules.
6 Such a restriction might seem uncharacteristically
autocratic in today's climate, where one of the features of the Communicative Approach is that students be given a choice of how they wish to express
themselves. It is our contention, however, that students have a true choice only if they have a variety
of linguistic forms at their disposal which they can
produce accurately. Without being restricted to
using a particular target form during a formfocused activity, students will often avoid producing the structure and, hence, never have an
opportunity to truly learn it.
7 An OS relative clause is one in which the subject of
the embedded sentence is replaced by a relative
pronoun because the subject is identical to an object
or objectlike noun in the predicate of the preceding
main clause. (For example: I like the book that he
wrote.) 0 s

266 Unit 11 E Language Skills/Grammar and Vocabulary