Chapter 3

Teaching: Input and Interaction

3. INTRODUCTION

We learned in chapter 2 how intake factors and intake processes interweave

and interact with each other in as yet undetermined ways to convert parts of
language input into learner intake. A crucial dimension of such a conver-
sion, particularly in the context of classroom L2 development, is the rela-
tionship between teaching strategies and learning outcomes. Several stud-
ies have been conducted to investigate the role and relevance of instruction
in the L2 classroom. One of the limitations of these studies is that they have
focused narrowly on grammatical instruction rather than on any wider as-
pect of language teaching. In fact, as learned in chapter 2, this limitation is
true not only of research related to teaching effectiveness but also research
in second-language acquisition in general and, therefore, we should always
keep in mind what Hatch (1978) said a quarter century ago about using re-
search findings for pedagogic purposes: Apply with caution.
Systematic investigation into the effect of language teaching (read:
grammar teaching) began as an offshoot of what came to be known as mor-
pheme studies (Dulay & Burt, 1974; Larsen-Freeman, 1976). These studies at-
tempted to assess whether, among other things, learning a language in
classroom settings is different from learning a language in naturalistic envi-
ronments. They revealed that the acquisition/accuracy order for various
grammatical morphemes like singular copula (’s/is), plural auxiliary (are),
possessive (’s), third person singular (-s), and so forth, is more or less the
same regardless of the learner’s L1 background, age, and learning environ-
ment (i.e., instructed or naturalistic). European researchers Wode (1976),

55
56 CHAPTER 3

Felix (1981) and their colleagues also found that the acquisition sequences
and strategies of L2 learners in classroom settings paralleled those followed
by L2 learners in naturalistic settings. Although these and other studies of a
similar kind dealt with only a handful of frequently occurring morphemes
among a multitude of grammatical structures that constitute language, they
hastily concluded that “the possibility of manipulating and controlling the
students’ verbal behavior in the classroom is in fact quite limited” (Felix,
1981, p. 109).
Such generalizations raised doubts about the effect of classroom instruc-
tion thereby prompting a very basic question: Does L2 instruction make any
difference at all? In order to explore this question, Long (1983) reviewed
11 studies on instructed L2 development conducted up to that point and
came out with ambiguous results. Six studies showed a positive effect of in-
struction, three showed minor or no effect, and two were unclear. Long,
however, concluded that formal instruction has positive effects on (a) L2
developmental processes, (b) the rate at which learners acquire the lan-
guage, and (c) their ultimate level of attainment. “Instruction is good for
you,” he declared rather encouragingly, “regardless of your proficiency
level, of the wider linguistic environment in which you receive it, and of the
type of test you are going to perform on” (1983, p. 379).
In spite of his encouraging conclusions, Long was concerned that the
11 studies available for his review were hardly the appropriate ones to
shed any collective light on the effect of instruction on L2 development.
The reason is threefold. First, the studies had very little research design in
common to put together to seek any common wisdom. Taken together,
they involved three types of learners (English as a second language, Eng-
lish as a foreign language, and Spanish as a second language), from three
different age groups (children, adolescents, and adults) with varying pro-
ficiency levels (beginning, intermediate, and advanced), learning their
target language in three different acquisition environments (rich, poor,
and mixed), responding to two different tests (discrete and integrative)
that sought to ascertain their learning outcomes. Secondly, as Long him-
self pointed out, most of the studies had failed to control for overall
amount of combined contact and instruction considering the fact that
they were conducted in environments where learners had access to the TL
through both formal and natural exposure. Finally, and perhaps most
importantly, the studies claiming to investigate the relationship between in-
struction and L2 development had bestowed only a scant attention on spe-
cific instructional strategies followed by classroom teachers who partici-
pated in the experiments. Besides, several teaching strategies were
clubbed together under generic terms thereby ignoring the possible ef-
fects of specific classroom strategies. Thus, the early studies on the effect
of instruction proved to be ineffectual, to say the least.
TEACHING: INPUT AND INTERACTION 57

In retrospect, it appears that we were asking the wrong question. Does in-
struction make a difference in L2 development? is as pointless and purposeless as
the question, Does nutrition make a difference in human growth? We can hardly
answer the first question in the negative unless we propose and defend the
untenable proposition that the human mind is untrainable. We all know
through experience that learners do learn at least a part of what is taught
and tested. The questions nutritionists normally ask are What kind of nutri-
tion makes a difference? and For who? Likewise, we should have asked ques-
tions such as What kind of instruction makes a difference? In what context? and
Using what method?
It comes as no surprise then that the initial inquiry into the effect of in-
struction has inevitably led to more focused studies with greater investiga-
tive rigor. Later studies (e.g., Donato & Adair-Hauck, 1992; Doughty, 1991;
Lightbown, 1992; Pica, 1987; Spada, 1987; Van Patten & Cadierno, 1993)
have not only sought to rectify some of the conceptual and methodological
flaws found in the early attempts but have also started focusing on the im-
pact of specific teaching strategies on learning-specified language items.
Most of these studies, however, still suffered from the earlier drawback of
dealing narrowly with grammatical instruction. Reviews of these and other
recent studies have shown that instruction does have a role to play (see
Doughty, 2003; Norris & Ortega, 2000). In her review of cases for and
against L2 instruction, Doughty (2003), for instance, concluded that “in-
struction is potentially effective, provided it is relevant to learners’ needs.
However, we will be forced to acknowledge that the evidence to date for ei-
ther absolute or relative effectiveness of L2 instruction is tenuous at best,
owing to improving, but still woefully inadequate, research methodology”
(p. 256).
Taken together, studies on L2 instruction suggest that proper instruc-
tional intervention at the proper time would be helpful for promoting de-
sired learning outcomes in the L2 classroom. This, of course, is not a star-
tling revelation because any language learning in a classroom context, as
against learning a language in a naturalistic setting, inevitably involves
some degree and some kind of intervention. We intervene by modifying the
content and style of language input, and we intervene by modifying the na-
ture and scope of interactional opportunities. Input modifications and
interactional activities, then, constitute the foundational structure of any
classroom learning and teaching operation.

3.1. INPUT MODIFICATIONS

It is generally agreed that language input has to be modified in order to

make it available and accessible to the learner. What has been the source of
disagreement is the type of modifications that should be brought about.
58 CHAPTER 3

The bone of contention centers around three strands of thought that can
be characterized as (a) form-based input modifications, (b) meaning-based
input modifications, and (c) form- and meaning-based input modifications.

3.1.1. Form-Based Input Modifications

Historically and until very recently, input modifications have almost always
been based on the formal (or structural) properties of the language,
whether they relate to grammatical forms or communicative functions. Lin-
guistic forms have been the driving force behind learning objectives, curric-
ulum design, materials production, classroom procedures, and testing tech-
niques. The essence of form-based input modifications, however, has not
remained constant. The changing norms can best be captured by positing a
product-oriented version and a process-oriented version of form-based in-
put modifications.
The product-oriented version of form-based input modifications treats
grammar as a product that can be analyzed, codified, and presented. It re-
lates to the characteristics of grammar teaching as propagated and prac-
ticed during the heyday of audiolingualism (see chap. 5, this volume, for
details). Within the audiolingual pedagogy, manipulating language input
meant selecting grammatical features, sequencing them in some fashion,
making them salient for the learner through a predominantly teacher-
centered, metalinguistic, decontextualized instruction involving explicit
pattern practice and explicit error correction. The learner was expected to
observe the grammatical input, examine it, analyze it, imitate it, practice it,
internalize it, use it. But, it became increasingly clear that confining the
learner to an exclusively product-oriented, form-based language input not
only distorted the nature of the target language exposed to the learner but
also decreased the learner’s potential to develop appropriate language
knowledge/ability. In short, the product-oriented version of form-based in-
put modifications turned out to be an extremist position.
The process-oriented version of form-based input modifications treats
grammar as a network of systems to be interacted with rather than an
objectified body of structures to be mastered. Instead of emphasizing mem-
ory, specific rules, and rule articulation, it focuses on understanding, gen-
eral principles, and operational experience. The input modifications advo-
cated here are still form-based but not based on teaching grammatical
structures per se but on creating what Rutherford (1987) called conscious-
ness raising. He explained that consciousness raising

is the means to an end, not the end itself. That is, whatever it is that is raised to
consciousness is not to be looked upon as an artifact or object of study to be
committed to memory by the learner and thence recalled by him whenever
TEACHING: INPUT AND INTERACTION 59

sentences have to be produced. Rather, what is raised to consciousness is not

the grammatical product but aspects of the grammatical process . . . (p. 104)

In the specific context of L2 learning and teaching, it refers to the deliber-

ate attempt to draw the learners’ attention to the formal processes of their
L2 in order to increase the degree of explicitness required to promote L2
development. Because consciousness is a loaded psychological term that can-
not be easily defined, Sharwood-Smith (1991) suggested a more verifiable
term, input enhancement, to refer to consciousness-raising activities. From a
pedagogic point of view, input enhancement serves the purpose of drawing
the learner’s explicit attention to grammatical features by such activities as
highlighting, underlining, rule-giving, and so forth.
The idea of grammatical process was recently expanded by Larsen-
Freeman (2000, 2003), who introduced the term, grammaring, to refer to
long-overlooked qualities of grammar such as that “it is a dynamic process
in which forms have meanings and uses in a rational, discursive, flexible,
interconnected, and open system” (Larsen-Freeman, 2003, p. 142). Gram-
maring is seen as the learner’s knowledge/ability to use grammatical struc-
tures accurately, meaningfully, and appropriately. Language input intro-
duced to the learner then should be modified in such a way as to make the
reason underlying a structure transparent. For example, Larsen-Freeman
suggests that when two different forms exist in a language, as in

There is a book on the table.

A book is on the table.

the underlying principle behind their variation in meaning or use must be

presented. As she explains

the meaning of these two sentences is more less the same, but the sentence
with there would be used to introduce new information in normal discourse.
The second sentence is much more limited in frequency and scope. One of its
functions is in giving stage directions to the director of a play, telling the di-
rector how to stage some scene in the play. While it may be difficult for stu-
dents to figure this difference out on their own, the principle will help them
learn to look for ways that particular grammar structures are distinctively
meaningful and/or appropriate. (Larsen-Freeman, 2000, p. 11)

Although the process-oriented, form-based input modifications appear

to have a greater intellectual appeal and instructional relevance than
strictly product-oriented, form-based input modifications, it must be re-
membered that proponents of both subscribe to similar, linguistically moti-
vated learning and teaching principles. That is, they believe that formal
properties of the language, both structures and relations, can be systemati-
60 CHAPTER 3

cally analyzed, selected, sequenced, and presented one by one to the

learner. They both believe that the learner will be able to put these discrete
items together in order to internalize the totality of language. Learners ex-
posed to such input modifications may be able to develop higher levels of
analysis of language as system but may not be able to understand the full im-
plications of communicative use. In other words, predominantly form-
based input modifications facilitate the development of linguistic knowl-
edge/ability but not necessarily pragmatic knowledge/ability both of
which, as we have seen in chapter 2, are required for successful language
communication. As a response to this predicament, it was suggested that
the focus be shifted from form to meaning.

3.1.2. Meaning-Based Input Modifications

A forceful articulation of the importance of meaning-based input modifica-

tions came a while ago from Newmark (1963/1970) who argued that “sys-
tematic attention to the grammatical form of utterances is neither a neces-
sary condition nor a sufficient one for successful language learning” and
that ”teaching particular utterances in contexts which provide meaning
and usability to learners is both sufficient . . . and necessary” (p. 217). Be-
cause these statements became very influential and are often misinter-
preted, it is important to recall the context in which these were made, and
also the caveat that accompanied them.
Newmark made these statements in a paper entitled “Grammatical The-
ory and the Teaching of English as a Foreign Language.” The grammatical
theory referred to here is Chomskyan transformational grammar, which
was newly proposed and widely discussed at that time. Emphasizing the in-
applicability of the theory of transformational grammar to language teach-
ing, Newmark asked language teachers to resist the “great temptations” to
write new language-teaching textbooks reflecting the “neat and precise”
grammatical analysis offered by transformational grammar. It is in this con-
text he suggested that, “we should liberate language teaching from gram-
matical theory, and should teach the natural use of language” (1963/1971,
p. 218). In a follow-up paper, Newmark (1966/1970) further clarified his
stand by saying that “the important point is that the study of grammar as such
is neither necessary nor sufficient for learning to use a language” (p. 226,
emphasis added). What he was objecting to is the study of grammar as such but
was in favor of “a limited kind of structural drill” so long as it is “embedded
in a meaningful context” (p. 226).
In spite of the context and the caveat, Newmark’s argument formed one
of the bases for an exclusively meaning-oriented input modification as ex-
emplified, for instance, in Krashen’s input hypothesis. To put it in a nut-
shell (see chap. 7, this volume, for details), the input hypothesis (Krashen,
TEACHING: INPUT AND INTERACTION 61

1982 and elsewhere) claimed that we acquire language in only one way: by
understanding messages, that is, by obtaining comprehensible input. Com-
prehensible input is defined as i + 1, structures that are a bit beyond the L2
learner’s current level of knowledge/ability. It is considered to contain all
the grammatical structures the acquirer is ready to acquire, in the right or-
der and right quantity, as long as enough comprehensible input of consis-
tently high quality is provided. Linguistic knowledge/ability is attained nec-
essarily and sufficiently as the result of mere exposure to instances of
comprehensible input, which can be provided through meaning-oriented
activities such as language games and problem-solving tasks. Form-based
language awareness does not play any direct role in L2 development. A sim-
ilar argument was made by Prabhu (1987), who stated that “the develop-
ment of competence in a second language requires not systematization of
language inputs or maximization of planned practice, but rather the cre-
ation of conditions in an effort to cope with communication” (p. 1).
Language-teaching programs that have systematically followed some of
the pedagogic features that later characterized the input hypothesis, and
for which we have a considerable body of research literature, are the Cana-
dian French immersion programs. These are public school programs in
which speakers of English (the majority language) study in French (the mi-
nority language). The learners have very little interaction with native speak-
ers of French other than their teachers, and exposure to French comes pri-
marily from teachers and instructional materials. Although the learners
seldom reach near native capability, they eventually emerge as competent
L2 speakers (Swain & Lapkin, 1982). According to Krashen (1984), immer-
sion “works” because it provides learners with a great deal of comprehensi-
ble input thereby proving that “subject-matter teaching is language teach-
ing” (pp. 61–62, emphasis in original).
Krashen’s enthusiastic endorsement notwithstanding, research based on
immersion as well as nonimmersion programs shows that exclusively mean-
ing-oriented input modifications do not lead to desired levels of grammati-
cal accuracy. Several studies (Lightbown 1992; Lightbown & Spada, 1990;
Schmidt, 1993; Van Patten 1990) have shown that even though learners ex-
posed to meaning-based input modifications speak fluently and confi-
dently, their speech is marked by numerous grammatical errors. In fact,
there is little evidence to show that successful grammar construction can
take place solely through meaning-based input modifications. Reviewing
more than two decades of research in French immersion classes, Swain
(1991) concluded that immersion students are able to understand much of
what they hear and read even at early grade levels, and that, although they
are well able to get their meaning across in their second language, even at
intermediate and higher grade levels, they often do so with nontargetlike
morphology and syntax. A probable reason is that language learners who
62 CHAPTER 3

are focusing on meaning may not have the processing space to attend to
form at the same time because of limitations on the number of cognitive
psychological operations learners can engage in. Whatever the reason, it is
clear that “learners do not very readily infer knowledge of the language sys-
tem from their communicative activities. The grammar, which they must
obviously acquire somehow as a necessary resource for use, proves elusive”
(Widdowson 1990, p. 161).
It turns out that it is not just grammatical knowledge/ability that proves
elusive; there may be problems in developing pragmatic knowledge/ability
as well. Citing examples from immersion studies, Swain (1991) argued that
“by focusing entirely on meaning, teachers frequently provide learners with
inconsistent and possibly random information about their target language
use” (p. 241). A specific example she cites to show how meaning-based in-
put modifications can be “functionally restricted” relates to the French pro-
nouns tu and vous, which carry information about both grammatical con-
cepts (singular, plural, or generic) and sociolinguistic use ( formal or informal).
An analysis of classroom input showed that the teachers used the pronouns
largely to denote their grammatical uses; there was scarcely any use of vous
in its sociolinguistic function marking politeness. This difference is consid-
ered the primary reason for the underuse of vous as a politeness marker by
immersion students.
Furthermore, even if the language input introduced by the teacher is
solely meaning-based, there is no way one can prevent learners from explic-
itly focusing on form, or vice versa. In other words, teachers may control in-
put availability in the classroom; they certainly do not control input accept-
ability. That belongs to the realm of the learner. Learners may be focusing
on both form and meaning regardless of teacher intention and interven-
tion. There are scholars who believe that a combination of form- and mean-
ing-based input is what is really needed.

3.1.3. Form- and Meaning-Based Input Modifications

Some of the carefully designed classroom-oriented experiments conducted

in the late 80s and early 90s (Doughty, 1991; Lightbown & Spada, 1990;
Spada, 1987; Van Patten & Cadierno, 1993) authenticated what the learn-
ers already seem to know, namely, focusing on form and meaning is more
beneficial than focusing on either one of them.
In a study on the development of oral communicative skills, Spada
(1987) investigated the relationships between instructional differences and
learning outcomes in three intermediate level classes of a communicatively
based ESL program. Class A received primarily form-based instruction,
Class B received both form- and meaning-based instruction, and Class C re-
ceived primarily meaning-based instruction. Her findings revealed that
TEACHING: INPUT AND INTERACTION 63

Class B registered a significant improvement, and Classes A and C did not

improve as much as Class B. She concluded that “neither form-based nor
meaning-based instruction in itself is sufficient, but rather, both are re-
quired” (p. 153). Her study reinforced her earlier finding that learners re-
quire opportunities for both form-focused and function-focused practice in
the development of particular skill areas, and if one or the other is lacking
they do not appear to benefit as much (Spada, 1986).
In a related study, Lightbown and Spada (1990) investigated the effects
of form-focused instruction and corrective feedback in communicative lan-
guage teaching. Their study was part of a long-term project and the data
came from more than a 1000 students in nearly 40 intensive ESL classes and
from over 200 students in regular ESL programs. The instructional strategy
consisted of meaning-based activities, opportunities for the negotiation of
meaning in group work, and the provision of comprehensible input. The
teachers who taught these classes differed from each other in terms of the
total amount of time they gave to form-focused activities. The researchers
analyzed the learners’ listening and reading comprehension as well as their
ability to speak. They found that form-based instruction within a communi-
cative context contributes to higher levels of language knowledge/ability.
Lightbown and Spada (1990) concluded that “accuracy, fluency, and over-
all communicative skills are probably best developed through instruction
that is primarily meaning-based but in which guidance is provided through
timely form-focused activities and correction in context” (p.443).
A similar conclusion was reached by Doughty (1991), who conducted an
experiment focusing on one grammatical subsystem of English (restrictive
relative clauses) with intermediate level international students from seven
different L1 backgrounds. They had very little knowledge of English rela-
tivization as revealed through a pilot test. They were randomly assigned to
one of three groups: two experimental groups (in addition to exposure to
relative clauses, the group was provided with an instructional treatment
aimed at improving their ability to relativize in English) and a control
group (in which they were exposed to relative clauses but received no in-
struction). Of the two experimental groups, one group (MOG) was given
meaning-oriented instruction along with the bringing to prominence of
the structural elements of relativization, and the other group (ROG) was
given exclusively rule-oriented instruction. The third group was called
COG (control group). The study revealed that compared to the control
group, both the MOG and ROG groups were equally effective with respect
to gain in relativization, but the MOG alone demonstrated substantial com-
prehension of the overall input. Doughty attributes the overall superior
performance of the MOG group to the successful combination of a focus
on meaning and the bringing to prominence of the linguistic properties of
relativization in the MOG treatment.
64 CHAPTER 3

The findings of the three experiments just outlined lead us to an inter-

esting proposition, namely, bringing linguistic properties to prominence
within the purview of a meaning-focused instructional strategy may change
the way language data are recognized by the learner as potential language
input, thus favorably shaping intake factors and intake processes (see chap.
2). Such a proposition has been put to test by Van Patten and Cadierno
(1993).
In a carefully designed study, Van Patten and Cadierno (1993) investi-
gated the relationship between instructional modifications and input proc-
essing, a term they use to refer to the process of converting input into in-
take. Based on a pretest, they randomly selected three groups of learners
studying Spanish as an L2 in the United States. The first group received
“traditional” instruction on object pronouns and word order, the second
received “processing” instruction on the same, and the third received no in-
struction at all on the targeted items. Traditional instruction involved pre-
senting the learners with explicit explanations concerning the form and
position of direct object pronouns within the sentence and then giving
them sustained practice, which moved the learners gradually from mechan-
ical drill to communicative drill. At all times, instruction focused on the
production of the targeted items by the learners, in other words, on their
output. In processing instruction, presentation was dominated by two types
of activities that forged form-meaning connections. One type had subjects
listening to or reading utterances and then demonstrating that they had
correctly assigned argument structure to the targeted items. The second
type of activity had subjects respond to the content of an utterance by
checking “agree” or “disagree.” At no point did processing instruction in-
volve the production of the targeted items by the learners. The results of
the experiment showed that unlike traditional instruction, processing in-
struction altered the way in which the learners recognized language input,
which in turn had an effect on the developing knowledge/ability of the
learners. Based on the results, Van Patten and Cadierno (1993) concluded
that “instruction is more beneficial when it is directed toward how learners
perceive and process input rather than when instruction is focused on hav-
ing learners practice the language via output” (p. 54).
In the context of helping learners actively engage form and meaning in a
principled way, Long (1991, 1996) proposed what is called focus on form
(not to be confused with form-focused input already discussed for which
Long uses the term focus on forms—note the plural. In order to avoid poten-
tial terminological confusion, I hereafter use its abbreviated version, FonF,
as suggested by Doughty & Williams, 1998). According to Long, FonF
“overtly draws students’ attention to linguistic elements as they arise inci-
dentally in lessons whose overriding focus is on meaning or communica-
tion” (Long, 1991, p. 46) and “consists of an occasional shift of attention to
TEACHING: INPUT AND INTERACTION 65

linguistic code features—by the teacher and/ or one or more students—

triggered by perceived problems with comprehension or production.
(Long & Robinson, 1998, p. 23). In other words, the learner’s attention to
linguistic features will be drawn explicitly if and only if it is necessitated by
communicative demand.
The input modification required for FonF places emphasis on designing
pedagogic tasks based on the future language needs of a particular group
of learners, tasks such as attending a job interview, making an airline reser-
vation, reading a restaurant menu or a journal abstract, writing a lab report,
or taking a driving test. For instance, learners may be given a task the solu-
tion of which requires them

to synthesize information on economic growth in Japan from two or more

written sources and use it to graph trends in imports and exports over a 10-
year period. Successful completions of the task involves them in reading (and
rereading) brief written summaries of sales trends for different sectors of the
Japanese economy, each of which uses such terms as rose, fell, grew, sank, plum-
meted, increased, decreased, declined, doubled, deteriorated, and exceeded. The fre-
quency of these lexical items in the input, due to their repeated use in the dif-
ferent passages, and/or their being underlined or italicized, makes them
more salient, and so increase the likelihood of their being noticed by stu-
dents. (Long & Robinson, 1998, pp. 24–25)

A task like this, as Doughty (2003) pointed out, helps learners integrate
forms and meaning, create their metalinguistic awareness, and increase
their noticing capacity all of which, as we discussed in chapter 2, promote
successful intake processing and ultimately language development.
An unmistakable lesson we learn from the aforementioned discussion is
that language should be presented to learners in such a way that they recog-
nize it as potential language input. We also learn that instruction should
help learners obtain language input in its full functional range, relevant
grammatical rules, and sociolinguistic norms in context along with helpful
corrective feedback. In other words, both form- and meaning-based input
modifications are essential for an effective L2 teaching program. Yet, just
the input, however modified, is not sufficient. What is additionally required
for learners to recognize and internalize form-meaning relationships is the
opportunity for meaningful interaction, and hence the importance of in-
teractional activities in classroom L2 learning and teaching.

3.2. INTERACTIONAL ACTIVITIES

Although the L2 literature presents several terms with attendant conceptual

ambiguities to refer to conversation in the classroom, the two that have been
widely used are interaction and negotiation. Both of them are used to mean
66 CHAPTER 3

something that is very different from their general usage involving intricate
sociolinguistic norms governing communication (see, e.g., the discussion
on Hymes’ SPEAKING acronym in chap. 1, this volume). The term interaction
or negotiation or negotiated interaction generally refers to conversational ex-
changes that arise when participants try to accommodate potential or actual
problems of understanding, using strategies such as comprehension checks
or clarification checks. Such an exercise is also perceived to promote the
learners’ processing capacity specifically by helping them with conscious no-
ticing required to convert input into intake (Gass, 1997; Long, 1996).
Characterizing such a definition of interaction as limited and limiting, I
have argued elsewhere (Kumaravadivelu, 2003a) that it is beneficial to iso-
late three interrelated dimensions of interaction and have discussed them
using, although it is a little bit of a stretch, Halliday’s macrofunctions of lan-
guage: textual, interpersonal, and ideational (see chap. 1, this volume, for
details). I have suggested that in the context of classroom communication,
we should actually talk about interaction as a textual activity, interaction as
an interpersonal activity, and interaction as an ideational activity. The first
refers to the linguistic realizations that create coherent written or spoken
texts that fit a particular interactional event, enabling L2 learners and their
interlocutors to understand the message as intended. Specifically, it focuses
on syntactic and semantic conversational signals, and its outcome is meas-
ured primarily in terms of linguistic knowledge/ability. The second refers
to the participants’ potential to establish and maintain social relationships
and have interpersonal encounters, and its outcome is measured in terms
of personal rapport created in the classroom. The third refers to an expres-
sion of one’s self-identity based on one’s experience of the real or imagi-
nary world in and outside the classroom. Specifically, it focuses on ideas
and emotions the participants bring with them, and its outcome is meas-
ured primarily in terms of pragmatic knowledge/ability. By introducing
such a tripartite division, I am not suggesting that the three dimensions are
equal or separate. Any successful interactional activity will mark the realiza-
tion of all three dimensions in varying degrees of sophistication. This divi-
sion is principally for ease of description and discussion. It is fair to say that
so far, L2 interactional research has focused largely on interaction as a tex-
tual activity, and to some extent on interaction as interpersonal activity. It
has almost completely ignored interaction as an ideational activity. Let us
briefly consider each of them.

3.2.1. Interaction as a Textual Activity

Most L2 interactional studies treat interaction primarily as a textual activity

in which learners and their interlocutors modify their speech phonologi-
cally, morphologically, lexically, and syntactically in order to maximize
TEACHING: INPUT AND INTERACTION 67

chances of mutual understanding, and minimize instances of communica-

tion breakdown. Such a seemingly excessive preoccupation with linguistic
aspects of interaction can best be understood in a historical perspective. A
major impetus for L2 interactional studies came from research on care-
taker talk conducted in the context of first-language acquisition. Empirical
studies carried out during the 1970s (R. Brown 1973; Snow 1972; Snow &
Ferguson 1977, and others) showed that the mother’s speech to the child
contained remarkably well-formed utterances characterized by a number of
formal adjustments in comparison to speech used in adult–adult conversa-
tions. The formal adjustments include: a lower mean length of utterances,
the use of sentences with a limited range of syntactic–semantic relations,
few subordinate and coordinate constructions, modified pitch, intonation
and rhythm, and a high level of redundancy.
Extending the concept of caretaker talk to L2 learners, researchers stud-
ied modified speech used by competent speakers of a language to outsiders
who were felt to have very limited or no knowledge/ability of it at all. This
modified speech has been referred to as foreigner talk. Ferguson (1975)
found that foreigner talk is very similar to caretaker talk. Specifically, he
found that foreigner talk is characterized by a slow rate of delivery, clear ar-
ticulation, pauses, emphatic stress, exaggerated pronunciation, paraphras-
ing, substitution of lexical items by synonyms, and omission, addition, and
replacement of syntactic features. Hopping from foreigner talk to teacher
talk was an easy and logical step. Not surprisingly then, teacher talk, that is,
the language a teacher uses to talk to L2 learners, was found to contain
characteristics of foreigner talk (Henzl, 1974). Further, it was found that
teacher talk increased in linguistic complexity with the increasing profi-
ciency level of the learners (Gaies, 1977).
Recognizing that L2 interactional studies so far had narrowly focused on
input, be it foreigner talk or teacher talk, and hence had overlooked “the
most important factor of all,” Hatch (1978) observed: “it is not enough to
look at input and to look at frequency; the important thing is to look at the
corpus as a whole and examine the interactions that take place within con-
versations to see how that interaction, itself, determines frequency of forms
and how it shows language functions evolving” (p. 403). The lead given by
Hatch has spawned several studies on the role of interaction resulting in a
substantial body of literature.
Foremost among the L2 interactionists is Long (1981, and elsewhere),
who makes a distinction between modified input and modified interaction.
The former involves a modification of language input that has short
phrases and sentences, fewer embeddings, and greater repetition of nouns
and verbs, whereas the latter involves a modification of the conversational
structure that has a considerable number of comprehension checks, confir-
mation checks, and clarification checks. As paraphrased by Allwright and
68 CHAPTER 3

Bailey (1991), a comprehension check is the speaker’s query of the inter-

locutors to see if they have understood what was said: Do you understand? or
Do you get what I’m saying? A confirmation check is the speaker’s query as to
whether or not the speaker’s (expressed) understanding of the interlocu-
tor’s meaning is correct: Oh, so are you saying you did live in London? A clarifi-
cation check is a request for further information or help in understanding
something the interlocutor has previously said: I don’t understand exactly.
What do you mean?
Long found that although modified input is unquestionably important,
it is participation in meaningful interaction made possible through modi-
fied interaction that significantly contributed to comprehension leading to
L2 development. Based on his work, he proposed a two-part hypothesis: (a)
interactional modifications geared to solving communication difficulties
help to make input comprehensible, and (2) comprehensible input pro-
motes L2 development. Subsequent research by Pica, Young, and Doughty
(1987) and others confirmed the first, but not the second, part of the hy-
pothesis. They found that learners who were exposed to linguistically un-
modified input with opportunities to negotiate meaning understood it
better than learners who were exposed to linguistically simplified version of
the input but were offered no opportunity for such negotiation.
Studies on interaction as a textual activity have clearly demonstrated that
interactional modifications help learners become aware of form-meaning
relationships. Several studies have questioned the claim that modified in-
put can be made comprehensible without any active participation on the
part of the learner. For instance, in a comparative study on the effects of in-
put modifications and of interactional modifications, Pica et al. (1987)
found that comprehension was assisted by the interactional modifications,
and that input modifications, even with reduced linguistic complexity, had
no such effect.
One does not have to look far to see the reasons for this. Input modifica-
tions, though crucial, do not by themselves offer opportunities for interac-
tion. They may make some of the structural–semantic features salient, but
they do not make structural–semantic relationships transparent. In other
words, input modifications may provide potentially acceptable input; but,
they do not help learners learn the relationship between form and mean-
ing in order to develop the necessary knowledge/ability to convey their in-
tended meaning in an interactive speech event. It is the learner’s inter-
actional efforts that make form-meaning relationships in the TL data
acceptable and internalizable. As Allwright and Bailey (1991) pointed out,
“it is the work required to negotiate that spurs language acquisition, rather
than the intended outcome of the work—comprehensible input” (p. 123).
Interactional modifications help learners focus on the meaningful use of
particular linguistic features, and practice the productive use of those fea-
TEACHING: INPUT AND INTERACTION 69

tures. They help learners stretch their limited linguistic repertoire, thereby
resulting in opportunities for further L2 development (for more details,
see Gass, 1997).
Although classroom interaction by definition includes learner produc-
tion, the role of learner output in L2 development was not given any seri-
ous consideration for a long time. The scope of interaction as a linguistic
activity has now been extended to include the effect of learner output, par-
ticularly after the emergence of two output-related hypotheses: the compre-
hensible output hypothesis (Swain, 1985) and the auto-input hypothesis
(Schmidt & Frota, 1986). Both these hypotheses emphasize the role played
by the learner’s output in shaping L2 development. They highlight the im-
portance of learner output produced in the process of meaningful interac-
tion as it provides the learner with the opportunity to form and test initial
hypotheses, and the opportunity to pay particular attention to the linguistic
means of communicative expression. A study by Pica, Holliday, Lewis, and
Morgenthaler (1989) found that comprehensible output was an outcome
of linguistic demands placed on the learner in the course of interaction.
Further research by Swain (1995) and others has confirmed the impor-
tance of output.
The precise role of interactional modifications in general has not been
sufficiently investigated (see Gass, 2003, for a recent review). However,
there seems to be a consensus among researchers that L2 learning environ-
ment must include opportunities for learners to engage in meaningful in-
teraction with competent speakers of their L2 if they are to discover the for-
mal and functional rules necessary for comprehension and production. As
the studies cited earlier show, what enables learners to move beyond their
current receptive and productive capacities when they need to understand
unfamiliar language input or when required to produce a comprehensible
message are opportunities to modify and restructure their interaction with
their interlocutors until mutual comprehension is reached.
That meaningful interaction is crucial for L2 development has been
widely recognized. There has not been adequate recognition, however, that
providing interactional opportunities means much more than providing
opportunities for an explicit focus on linguistic features or for a possible
form-meaning relationships embedded in the input data. Studies that ap-
proach interaction primarily as a textual activity can offer only a limited
perspective on the role of interaction in L2 development, for they treat
interactional modifications as no more than conversational adjustments.
Clearly, interaction is much broader a construct than that. It entails, mini-
mally, a spectrum of linguistic, social, and cultural constructs that create
the very context of language communication. Therefore, in order to facili-
tate an effective interplay of various intake factors and intake processes dis-
cussed in the previous chapter, we may have to go beyond the narrow con-
70 CHAPTER 3

fines of interaction as a textual activity, and consider the role of interaction

as an interpersonal activity and also interaction as an ideational activity,
among other yet unknown possibilities.

3.2.2. Interaction as an Interpersonal Activity

Unlike interaction as a textual activity that deals with conversational adjust-

ments, interaction as an interpersonal activity deals with interpersonal com-
munication. Classroom community is a minisociety nested within a larger
society. It has its own rules, regulations, and role relationships. Interaction
in such a minisociety is essentially a social process involving, as Breen
(1985) pointed out, all its participants in verbal and nonverbal interaction
that exists on a continuum from ritualized, predictable, phatic communion
to dynamic, unpredictable, diversely interpreted communication, just as in
any social interaction. Classroom community presents different contexts
for different participants who bring different social realities with them. It
also represents a tension between the internal world of the individual and
the social world of the group. This tension requires individuals to adapt
their learning process to the sociopsychological resources of the group, just
as the group’s psychic and social process unfolds from the individual contri-
butions of a learner.
Interaction as an interpersonal activity, therefore, has the potential to
create a conducive atmosphere in which the other two interactional activi-
ties—textual and ideational—can flourish. Such a potential has not been
adequately explored, much less exploited. Studies conducted by Wong-
Fillmore (e.g., 1989) reveal that social processes are as important as cogni-
tive processes for successful L2 development. As we have seen in chapter 2,
social processes are steps by which both the learners and competent speak-
ers of the TL create and shape appropriate social settings in which it is pos-
sible and desirable to communicate by means of the TL. In a research
study, Donato and Adair-Hauck (1992) showed how groups or dyads en-
gaged in social interactions both in and outside the classroom foster the
formation of linguistic awareness in learners. Taking a Vygotskyan perspec-
tive to social processes of language learning, they argue for an interactional
approach in which social discourse is central to the teaching–learning rela-
tionship.
Vygotskyan sociocultural theory provides a richer and deeper interpreta-
tion of the role of interaction in the language classroom (Hall, 2002;
Lantoff, 2000). It focuses on the construction of interpersonal interactions
where participants actively and dynamically negotiate not just textual mean-
ing, but also their social relationships. Such an approach treats interaction
as a social practice that shapes and reshapes language learning. Thus, as
Ellis (1999) explained, “socio-cultural theory has the greater potential as it
TEACHING: INPUT AND INTERACTION 71

emphasizes the collaborative nature in meaning making in discourse in

general, not just in exchanges where communication breakdown occurs”
(p. 224).
In fact, at the pedagogic core of interaction as an interpersonal activity
are opportunities for increased learner–learner interaction and greater
topic control on the part of the learner. Learner–learner interaction, other-
wise known in the L2 literature as nonnative speaker/nonnative speaker
(NNS/NNS) interaction, was initially thought to provide what is called
“junky” input data, which can hardly help on successful L2 learning. How-
ever, Yule and Gregory (1989), for instance, found “sufficient evidence to
suggest that the benefits of modified interaction, in terms of creating more
comprehensible input, can actually be obtained in a situation which does
not involve native speaking interlocutors” (p. 42). Similarly other studies
on classroom interactional analysis demonstrate that NNS/NNS interactive
discourse is equally beneficial in promoting L2 comprehension and pro-
duction (see Gass, 1997). According to these studies, NNS/NNS partners
produce more and frequent instances of interactional modifications, and
employ more communication strategies than do NS/NNS partners thereby
enhancing their chances of L2 comprehension.
Closely linked to the opportunity made available for learner–learner in-
teraction is the flexibility given to learners in nominating topics for discus-
sion in class. During the early part of interactional research, Hatch (1978)
reported that giving the learners the freedom to nominate topics provided
an effective basis for interactional opportunities. Although not enough
work has been done on the effect of learner topic control, a study by
Slimani (1989) found that learners benefited more from self- and peer-
nominated topics than from teacher nominated topics. Reflecting on
learner-topic control, Ellis (1992) rightly observes,

Having control over the topic is also one way of ensuring that the linguistic
complexity of the input is tailored to the learner’s own level. Better opportu-
nities for negotiating meaning when a communication problem arises are
likely to occur. Topic control may also stimulate more extensive and more
complex production on the part of the learner (p. 177)

3.2.3. Interaction as an Ideational Activity

Both interaction as textual and interaction as interpersonal activities can
provide only a limited perspective because they do not take into account
the social, cultural, political, and historical processes and practices that
shape language learning and teaching. In other words, they both fail to rec-
ognize language as ideology (cf. chap. 1, this volume). Language, as Wee-
den (1997) so aptly stated, “is the place where actual and possible forms of
social organization and their likely social and political consequences are de-
72 CHAPTER 3

fined and contested. Yet it is also the place where our sense of ourselves,
our subjectivity, is constructed” (p. 21). Thus, language is not simply a net-
work of interconnected linguistic systems; rather, it is a web of interlinked
sociopolitical and historical factors that shape one’s identity and voice. In
such a context, the development of the ability to speak one’s mind and “the
ability to impose reception” (Bourdieu, 1991) are of paramount impor-
tance. It is, therefore, no longer sufficient if interactional modifications
provide the learners only with the opportunity to fix communication break-
downs or to foster personal relationships in class. They must also provide
them with some of the tools necessary for identity formation and social
transformation.
Nobody emphasized this critical nature of education more and with
greater conviction than critical pedagogists such as Giroux, Shor, Simon,
and others who, influenced by the pioneering thoughts of Paulo Freire,
looked at the classroom as an ideological site—a site that is socially con-
structed, politically motivated, and historically determined. Therefore, crit-
ical pedagogy has to empower classroom participants “to critically appropri-
ate forms of knowledge outside of their immediate experience, to envisage
versions of a world which is ‘not yet’ in order to alter the grounds on which
life is lived” (Simon, 1988, p. 2). Such a pedagogy would take seriously the
sociopolitical, historical conditions that create the cultural forms and inter-
ested knowledge that give meaning to the lives of teachers and learners. “In
one sense, this points to the need to develop theories, forms of knowledge,
and social practices that work with the experiences that people bring to the
pedagogical setting” (Giroux, 1988, p.134).
Critical pedagogists call for an “empowering education” that relates
“personal growth to public life by developing strong skills, academic knowl-
edge, habits of inquiry, and critical curiosity about society, power, inequal-
ity, and change” (Shor 1992, p. 15); and one that helps students explore the
subject matter in its sociopolitical, historical contexts with critical themes
integrated into student language and experience. They consider contem-
porary language education “as somewhat bizarre in that it legitimates and
limits language issues as technical and developmental” (Giroux & Simon,
1988, p. 131) and believe that language education must be “viewed as a
form of learning that not only instructs students into ways of ‘naming’ the
world but also introduces them to particular social relations” (Giroux & Si-
mon, 1988, p. 131).
In the same vein, critical linguists argue that “all representation is medi-
ated, moulded by the value-systems that are ingrained in the medium (lan-
guage in this case) used for representation; it challenges common sense by
pointing out that something could have been represented in some other
way, with a very different significance” (Fowler, 1996, p. 4). Saying that ide-
ology and power that constitute dominant discourses are hidden from ordi-
TEACHING: INPUT AND INTERACTION 73

nary people, critical linguists seek to make them visible by engaging in a

type of critical discourse analysis that “is more issue-oriented than theory-
oriented” (van Dijk, 1997, p. 22). By doing so, they hope to shed light on
the way power relations work within the society. They thus move from the
local to the global displaying “how discourse cumulatively contributes to
the reproduction of macro structures . . .” (Fairclough, 1995, p. 42).
As can be expected, critical linguists pointedly emphasize the role of crit-
ical language awareness in developing sociopolitical consciousness. Fair-
clough, in particular, believes that critical language awareness “can lead to
reflexive analysis of practices of domination implicit in the transmission
and learning of academic discourse, and the engagement of learners in the
struggle to contest and change such practices” (Fairclough, 1995, p. 222).
He further points out that language learners can learn to contest practices
of domination only if the relationship between language and power is made
explicit to them.
Pointing out that researchers in L2 interactional analysis have shied away
from any serious engagement with the ideological forces acting upon class-
room discourse, I have proposed what is called critical classroom discourse
analysis (CCDA; Kumaravadivelu, 1999a). The primary function of such an
analysis is to play a reflective role, enabling teachers to reflect on and to
cope with sociocultural and sociopolitical structures that directly or indi-
rectly shape the character and content of classroom interaction. I have ar-
gued that

language teachers can ill afford to ignore the sociocultural reality that influ-
ences identity formation in and outside the classroom nor can they afford to
separate learners’ linguistic needs and wants from their sociocultural needs
and wants. Negotiation of discourse meaning and its analysis should not be
confined to the acquisitional aspects of input and interaction, or to the in-
structional imperatives of form/ function focused language learning activities
or to the conversational routines of turn-taking and turn-giving sequences; in-
stead, they should also take into account discourse participants’ complex and
competing expectations and beliefs, identities and voices, fears and anxieties.
(Kumaravadivelu, 1999a, p. 472)

Drawing from the CCDA perspective, I suggest that interaction as an

ideational activity must necessarily address questions such as:

· If classroom interaction is socially constructed, politically motivated,

and historically determined, what are the ways in which we can study
and understand the impact of these forces on interactional modifica-
tions?
· If discourse participants bring to the classroom their racialized, strati-
fied, and gendered experiences, how can we identify the way(s) in
74 CHAPTER 3

which these experiences motivate the style and substance of classroom

interaction?
· If the objective of language education should not be merely to facili-
tate effective language use but also to promote critical engagement
among discourse participants, then how can we analyze and assess the
extent to which critical engagement is facilitated in the classroom?
· If the learner’s voice has to be recognized and respected, how might
their personal purposes, attitudes, and preferred ways of doing things
be reconciled with interactional rules and regulations, and instruc-
tional aims and objectives?
· If negotiation of discourse meaning is not confined to the acqui-
sitional aspects of input and interaction, but include expectations and
beliefs, identities, and voices, fears and anxieties of the participants,
how might such a comprehensive treatment help or hinder the proper
management of classroom interaction?
· If classroom discourse lends itself to multiple perspectives depending
on discourse participants’ preconceived notions of learning, teaching,
and learning outcomes, how can we identify and understand possible
mismatches between intentions and interpretations of classroom aims
and events?

Clearly, investigations of these and other related questions will provide ad-
ditional insights necessary to determine the nature and scope of interaction
as an ideational activity.
An increasing number of scholars in L2 learning and teaching have ex-
pressed similar critical thoughts about power and inequality in L2 educa-
tion as well. For example, Norton (2000) introduced the concept of invest-
ment, which presupposes that when language learners interact, they are not
only exchanging information but “are constantly organizing and reorganiz-
ing a sense of who they are and how they relate to the social world. Thus an
investment in the target language is also an investment in a learner’s own
identity, an identity which is constantly changing across time and space”
(pp. 10–11). Similarly, Benesch (2001), demonstrated how “all teaching is
ideological, whether or not the politics are acknowledged” (p. 46), and has
shown us how teaching English for academic purposes can usefully address
students’ multiple identities by engaging them in decisions affecting their
lives in and out of school. Hall (2002) argued for a teaching agenda that is
embedded in a sociohistorical and/or sociopolitical authority. Johnston
(2003) called for a particular way of seeing the language classroom and has
sought “to reveal the value-laden nature of our work in the language class-
room and to provide tools for analyzing that work” (p. 5). A common
thread that runs through all these works is an unfailing emphasis on inter-
action as an ideational activity.
TEACHING: INPUT AND INTERACTION 75

To sum up this section on interactional activities, if interaction as a textual

activity focuses on formal concepts, and interaction as an interpersonal activ-
ity focuses on social context, then interaction as an ideational activity may be
said to focus on ideological content. If the first enables learners to modify
conversational signals, the second encourages them to initiate interactional
topics, the third empowers them to construct their individual identity. If first
measures quality of interaction in terms of gains in linguistic knowledge, the
second measures it in terms of gains in sociocultural knowledge. The three
types of interaction may be said to produce three types of discourse: (a) inter-
action as a textual activity produces instructional discourse resulting in better
conversational understanding; (b) interaction as an interpersonal activity
produces informational discourse resulting in superior social communica-
tion; and (c) interaction as an ideational activity produces ideological dis-
course resulting in greater sociopolitical consciousness. These three types of
activities, however, should not be viewed as hierarchical, that is, they should
not be associated with the traditional levels of proficiency—beginning, inter-
mediate, and advanced. From a language-acquisitional point of view, they
make it easier for learners of various levels to notice potential language in-
put, and recognize syntactic–semantic relationships embedded in the input,
thereby maximizing their learning potential.
Instructional design that deals with the selection and sequencing of lan-
guage content in order to maximize the interplay between input and inter-
action on one hand, and the learner and the learning process on the other
hand, is yet another important piece of the pedagogic puzzle. In the next
section, I turn to the design issues under the general rubric: content specifi-
cations.

3.3. CONTENT SPECIFICATIONS

One of the essential components of any language teaching program is sylla-

bus or curriculum, which specifies the what or the content of language learn-
ing and teaching. The two terms are often used interchangeably although
they may indicate a hierarchical relationship where curriculum refers
broadly to all aspects of language policy, language planning, teaching
methods, and evaluation measures, whereas syllabus relates narrowly to the
specification of content and the sequencing of what is to be taught. This
section is limited to syllabus as a content-specifier.

3.3.1. Syllabus Characteristics

A well-designed language teaching syllabus seeks mainly (a) to clarify the

aims and objectives of learning and teaching, and (b) to indicate the class-
room procedures the teacher may wish to follow. More specifically, any
76 CHAPTER 3

syllabus, according to Breen (2001, p. 151), should ideally provide the fol-
lowing:

· A clear framework of knowledge and capabilities selected to be appro-

priate to overall aims;
· continuity and a sense of direction in classroom work for teacher and
students;
· a record for other teachers of what has been covered in the course;
· a basis for evaluating students’ progress;
· a basis for evaluating the appropriateness of the course in relation to
overall aims and student needs, identified both before and during the
course;
· content appropriate to the broader language curriculum, the particu-
lar class of learners, and the educational situation and wider society in
which the course is located.

Of course, the assumption behind this ideal list of syllabus objectives is that
they will enable teaching to become more organized and more effective. In
that sense, a syllabus is more a teaching organizer than a learning indicator,
although a well-conceived and well-constructed syllabus is supposed to re-
late as closely as possible to learning processes.
But to expect any close connection between teaching design and learn-
ing device is to ignore the role of learner intake factors on intake processes
that we discussed in chapter 2. It is precisely for this reason Corder (1967)
talked about the notion of a “built-in-syllabus” that learners themselves con-
struct based on the language content presented to them and in conjunc-
tion with intake factors and processes. As Corder rightly asserted, the
learner syllabus is organic rather than linear, that is, learners appear to
learn several items simultaneously rather than sequentially retaining some,
rejecting others and reframing certain others. What is therefore needed is a
psycholinguistic basis for syllabus construction.
A well-known work that attempted to determine a possible set of
psycholinguistically valid criteria for syllabus construction was reported by
Manfred Pienemann and his colleagues. In a series of empirical studies,
Pienemann (1984, 1987) investigated the acquisitional sequence of Ger-
man word order rules:

Stage 1: X = canonical order

Romance learners of German as a Second Language (GSL) start out with

a subject–verb–object order as their initial hypothesis about German word
order, for example, die kinder spielen mit ball (‘the children play with the ball’).
TEACHING: INPUT AND INTERACTION 77

Stage 2: X + 1 adverb-preposing

For example, da kinder spielen (‘there children play’). This preposing rule
is optional in German. But once this rules is applied, Standard German re-
quires a word order like ‘there play children’ (i.e., inversion).

Stage 3: X + 2 = verb separation

For example, alle kinder muß die pause machen (‘all children must the
break have’). Before the verb separation is acquired, the word order in the
interlanguage is the same as in sentences with main verbs only (cf. the Eng-
lish equivalent—all children must have a break). Verb separation is obligatory
in Standard German.

Stage 4: X + 3 = inversion

For example, dann hat sie wieder die knoch gebringt (‘then has she again the
bone bringed’). In Standard German, subject and inflected verbal element
have to be inverted after preposing of elements.
From a group of Italian children learning German as a second language
in a naturalistic environment, Pienemann selected 10 who were either at
Stage 2 or Stage 3 in their L2 development. The subjects were given class-
room instruction for 2 weeks on the structure from Stage 4, that is, inver-
sion. When they were tested for the development of the newly instructed
structure, Pienemann found that children who were at Stage 3 progressed
to Stage 4, but children who were at Stage 2 remained at the same stage.
The study, he surmised, demonstrated that the relevant acquisitional stages
are interrelated in such a way that at each stage, the processing prerequi-
sites for the following stage are developed.
Based on his findings, Pienemann proposed what he called a learnability/
teachability hypothesis. The learnability hypothesis states that learners can
benefit from classroom instruction only when they are psycholinguistically
ready for it. The learnability of a structure in turn constrains the effective-
ness of teaching, which is the teachability hypothesis. The teachability hy-
pothesis predicts that instruction can only promote language acquisition if
the interlanguage of the L2 learner is close to the point when the structure
to be taught is acquired in the natural setting so that sufficient processing
prerequisites are developed.
Notice that the teachability hypothesis does not claim that teaching has no
influence whatsoever on L2 development. Rather, it maintains that the influ-
ence of teaching is restricted to the learning items for which the learner is
ready to process. Pienemann argued that, provided the learner is at the ap-
propriate acquisitional stage, instruction can improve acquisition with re-
spect to (a) the speed of acquisition, (b) the frequency of rule application,
78 CHAPTER 3

and (c) the different linguistic contexts in which the rule has to be applied.
From his findings, Pienemann derived two general tenets for L2 teaching:

The principles of L2 development are not only a more reliable background

for psycholinguistically plausible simple–complex criteria in material grading
than the present intuitive procedures, but they are a necessary background
for grading, since formal L2 learning is subject to a set of learning principles
which are shared by formal and natural L2 developments. Thus, teaching is
only possible within the margin determined by these principles. As a conse-
quence, any learning task which contradicts these principles is not-learnable;
it would ask too much of the learner. (Pienamann, 1984, pp. 40–41)

The learnability/teachability hypothesis as an idea makes eminent sense

and has pointed toward a fruitful line of research (see Pienamann, 2003,
for a recent review of his and related works). However, its validity and its ap-
plicability have been questioned because of the small size of the sample and
also because of practical problems, like identifying the learners’ current state
of grammar. Besides, further research by others (e.g., Lightbown, 1985)
demonstrated that classroom learners develop their language in a sequence
that has no bearing on the sequence introduced by the teacher. The general
consensus now is that we just do not have adequate knowledge of the
learner’s language-processing capacity in order to coordinate the teaching
sequence with learning sequence.
In spite of the advances made in psycholinguistic research, our rationale
for selecting and grading language input presented to the learner is no
more objective today than it was more than a quarter century ago when
Mackay (1965) discussed the highly subjective notions of “difficulty” and
“complexity.” Pointing out that selection is an “inherent” characteristic of
any language teaching enterprise because, “it is impossible to teach the
whole of language,” Mackey (1965) identified three major criteria for selec-
tion: frequency, range, and availability. Frequency refers to the items that oc-
cur the most often in the linguistic input that the learners are likely to en-
counter. It is, therefore, tied to the linguistic needs and wants of the
learners. Range, on the other hand, is the spread of an item across texts or
contexts. In other words, an item that is found and used in several commu-
nicative contexts is more important than the one that is confined to one or
two contexts. Although frequency of an item answers the question how of-
ten it occurs, range answers the questions where it is used, by whom, and for
what purposes. Availability relates to the degree to which an item is neces-
sary and appropriate, and it also corresponds to the readiness with which it
is remembered and used.
Gradation deals with sequencing (which comes before which) and
grouping (what goes with what) of linguistic items. According to Kelly
(1969), syllabus designers have historically used three basic principles for
TEACHING: INPUT AND INTERACTION 79

determining the sequencing of linguistic input: complexity, regularity, and

productivity. The first principle suggests a movement from the easy to the
difficult, the second from the regular to the irregular, and the third from
the more useful to the less useful. Unlike sequencing, grouping is con-
cerned with the systems of a language, and its structures (Mackey, 1965).
Grouping attempts to answer the question: What sounds, words, phrases, or
grammatical structures can be grouped and taught together? For instance,
the simple present (habitual) may be grouped with words like usually, often,
and every, as in I go to the park every weekend. Similarly, words may be grouped
together by association (chair, table, furniture, seat, sit, etc.).
The putting together of the selected and graded language input is gen-
erally governed by the overall theoretical stance adopted by the syllabus de-
signer. Once again, the L2 literature presents a plethora of syllabuses as re-
flected in labels such as the structural syllabus, the notional-functional
syllabus, the task-based syllabus, the discourse syllabus, the skill-based sylla-
bus, the content-based syllabus, the process-syllabus, the procedural sylla-
bus, and so forth. Although one can discern subtle and sometimes signifi-
cant variations among these in terms of content as well as method of
teaching, there are certainly overlapping features among them. A fruitful
way of understanding the basic philosophy governing these types of syllabus
is to put them into broad classifications.

3.3.2. Syllabus Classifications

Nearly a quarter century ago, Wilkins (1976) proposed two broad classifica-
tions of syllabus: synthetic syllabus and analytic syllabus. The underlying as-
sumption behind the synthetic syllabus is that a language system can be (a)
analyzed into its smaller units of grammatical structures, lexical items, or
functional categories; (b) classified in some manageable and useful way;
and (c) presented to the learner one by one for their understanding and as-
similation. The learners then are expected to synthesize all the separate ele-
ments in order to get the totality of the language. Because the synthesis is
done by the learner, the syllabus is dubbed synthetic. The language-cen-
tered as well as learner-centered methods discussed in chapter 5 and chap-
ter 6 follow the synthetic syllabus. As we see in much detail in those chap-
ters, language-centered pedagogists devised suitable classroom procedures
for teachers to present, and help learners synthesize, discrete items of
grammar and vocabulary while learner-centered pedagogists did the same,
adding notional and functional categories to the linguistic items.
In the analytic syllabus, the language input is presented to the learner, not
piece by piece, but in fairly large chunks. These chunks will not have any spe-
cific linguistic focus; instead, they will bring the learner’s attention to the
communicative features of the language. They are connected texts in the
80 CHAPTER 3

form of stories, games, problems, tasks, and so forth. It is the responsibility of

the learner to analyze the connected texts into its smaller constituent ele-
ments, hence the term, analytic. Learning-centered methods discussed in
chapter 7 adhere to the analytic approach to syllabus construction.
It is not the purpose of this chapter to discuss in detail how these syllabus
types are linked to other aspects of language teaching such as teaching
strategies, textbook production, and evaluation measures. These will be ex-
plained with examples as we discuss different categories of method in Part
II of the volume.

3.4. CONCLUSION

In this chapter, I focused on various aspects of input, interaction, and sylla-

bus design as they impact on classroom instruction. In spite of the impres-
sive knowledge we have gained on the nature and relevance of input and
interactional modifications, we have only a limited understanding of their
role in L2 learning and teaching. A primary reason is that, as mentioned
earlier, studies on classroom instruction have focused generally and nar-
rowly on the impact of grammatical instruction rather than on the intricate
and intractable issue of the interplay between input and interaction on one
hand and between them and intake factors and intake processes on the
other hand. The fact that research on instructional modifications has not
substantively addressed this crucial relationship should have a sobering in-
fluence on our readiness to draw implications for pedagogic purposes.
But still, applied linguists are left with no option but to make use of the
still developing knowledge for drawing useful and useable ideas for lan-
guage teaching. According to Corder (1984),

There are those who believe that second language acquisition research is still
at such a preliminary stage that it is premature to base any proposals for lan-
guage teaching upon it yet. There are others, among whom I count myself,
who believe that it is the task of the applied linguist to make practical use of
whatever knowledge is available at the time. We cannot constantly be waiting
to see what is around the next corner. (p. 58)

Indeed, without waiting to see what is around the next corner, applied
linguists have, from time to time, readily conceived and constructed a suc-
cession of language-teaching methods based on insights from whatever re-
search findings that were available to them. In the same way, I attempt to
use the current state of knowledge to describe and evaluate their successes
and failures in order to see what we can learn from them. More specifically,
I use the features of language, learning, and teaching discussed so far to
take a close and critical look at major categories of language teaching meth-
ods. With that objective in mind, let us turn to Part II.