Discourse

[Cook]

1. What is Discourse?

1.1Discourse and the Sentence:

We have two different kinds of language as potential objects for

study:
One abstracted in order to teach a lang or literacy, or to study how
the rules of language work
And another which has been used to communicate something, and it
´s felt to be coherent (and may or may not correspond to a correct
sentence or a series of correct sentences).
The latter kind is language in use, for communication, is called
Discourse; and the search for what gives discourse coherence is
Discourse Analysis.
The two approaches are not mutually exclusive. Discourse may be
composed of one or more well-formed grammatical sentences but it
doesn´t have to be. It can have grammatical `mistakes´ in it, and
often does. Discourse treats the rules of grammar as a resource,
conforming to them when it needs to, but departing from them when
it does not.
What matters is not its conformity to rules, but the fact that it
communicates and is recognized by its receivers as coherent. There is
a degree of subjectivity in identifying a stretch of language as
discourse, yet in practice we find that discourse is usually perceived
as such by groups, rather than individuals.

1.2Grammar withing and beyond that sentence:

Grammar doesn´t stop with a full stop but reaches over it. There are
also rules which limit what kind of sentence can follow another. In the
same way that there are rules within sentences, limiting which words
can follow others, so there might also be rules within discourses,
limiting which sentences can follow another one.
So we now have two possible answers to the problem of how we
recognize a stretch of language as unified and meaningful. One is that
we employ lang rules of the type studied by grammarians and taught
in most language textbooks, and that these rules operate between
sentences as well as within them. The other is that we employ
knowledge – of the world, of the speaker, of social convention, of what
is going on around us as we read or listen – in order to make sense of
the language we are encountering. Coherence is created by factors
outside language.

1.3Language in and out of context:

When we receive a linguistic message, we pay attention to many
other factors apart from the language itself. If we are face to face with
the person sending the message, then we notice what they are doing
with their face, eyes, and body while speaking. In a spoken message
we notice the quality of the voice as well. These are called
paralinguistic features.

We are also influenced by the situation in which we receive messages,

by our cultural and social relationship with the aprticipants, by what
we know and what we assume the sender knows. These factors take
us beyond the study of language, in a narrow sense, and force us to
look at other areas of inquiry – the mind, the body, society, the
physical world – in fact, at everything. There are good arguments for
limiting a field of study to make it manageable; but it is also true to
say that the answer to the question of what gives discourse its unity
may be impossible to give without considering the world at large: the
Context.
In linguistics, specially in the English-speaking world bet the 1930s
and 1960s there have been several schools ot thought which believe
that context - this knowledge outside the world outside lang which
we use to interpret it – should be ruled out of language analysis as far
as possible. In this way, it is believe, linguists will be able to make
discoveries abt the language itself, and its system of rules which
exists quite independently of particular circumstances. We may
validly characterize it as sentence linguistics, because it confines its
inquiries to what happens within sentences. Sentence linguists follow
one of two procedures: they eitherinvent their examples for analysis,
using their own intuitive knowledge as native speakers (their linguistic
competence) as a yardstick, or they take language which people have
actually used and remove all the features which tey believe to be
irrelevant to their purposes. This process of eliminating the unique
combination of circumstances in which language happens is
technically known as Idealization.
Yet for the discourse analyst it may be exactly these transient and
variable features which enable us to understand the meaning of what
is said, and the reason why the order of sentences proceeds in the
way that it does. The langauge learner needs to be able to handle
language which is not idealized – language in use. The lang teacher
needs, therefore, to decide on the extent to which idealized language
may help the development of this ability.
We have then, two approaches to language: sentence linguistics and
discourse analysis. Both have an invaluable contribution to make to
the understanding of language, and both ultimately need each other.
We cannot communicate with only the rules of semantics and
grammar, so we just as surely cannot communicate very well without
them.
Sentence Linguistics Data Discourse Analysis
Data

Isolated sentences Any stretch of lang felt to

be unified
Grammatically well-formed Achieving meaning
Without context In context
Invented or idealized Observed

1.4The origins of discourse analysis:

The first known students of language in the Western tradition, the

scholars of Greece and Rome, were aware of these different
approaches too, and divided grammar from rhetoric, the former being
concerned with the rules of languge as an isolated object, the latter
with how to do things with words, to achieve effects, and
communicate successfully with people in particular contexts.
Ironically, some schools of discourse analysis – often thought of as
one of the newest disciplines of language study – employ terms from
classical rhetoric, one of the oldest. And there have always been,
throughout history, studies of language in context. In North America,
in the early decades of this century, exciting work was conducted by
people who were at once both anthropologists and linguists, often
involved in research into the languages and societies of the native
Americans (Indians). In Britain J.R. Firth, saw language not as an
autonomous system, but as part of a culture, which is in turn
responsive to the environment.
There are many other disciplines – which often examine their object of
study – through language, and are thus carrying out their own
discourse analysis. Perhaps the most useful distinction is to think of
other disciplilnes as studying something else through discourse;
whereas discourse analysis has discourse as its prime object of study,
and though it may take excursions into many different fields, must
always be careful to return to the main concern.
Ironically, it was a sentence linguist who both coined the term
`discourse analysis´ and initiated a search for lang rules which would
explain how sentences were connected withing a text by a kind of
extended grammar. This was Zellig Harris. In 1952, in an article
entitled `Discourse Analysis´, he analysed an advertisement for hair
tonic, but his conclusions are extremely interesting. At the beginning
of the article he observed that there were two possible directions for
discourse analysis. One was `continuing descriptive Linguistics
beyond the limits of a single sentence at a time´. This was what he
aimed to achieve. The other was `correlating culture and language
(i.e. non-linguistic and linguistic behaviour)´. But having weighed up
the two options, at the end of the article, he concluded: `... in every
language it turns out that almost all the results lie within a relatively
short stretch which we may call the sentence... Only rarely can we
state restrictions across sentences´.
If we are to find an answer to the problem of what gives stretches of
language unitiy and meaning, we must look beyond the formal rules
operating with sentences, and consider the people who use language,
and the world in which it happens as well. Yet before we do so, it
would be as well to see just how far formal, purely lingusitic rules can
go in accounting for the way one sentence succedes another.

2. Formal Links

2.1 Formal and contextual links:

In order to account for discourse, we need to look at features outside

the language. This facts enable us to construct stretches of lang as
discourse, as having a meaning and a unity for us. The way we
recognize correct and incorrect sentences is different. We can do this
through our knowledge of grammar without reference to outside
facts.
We can describe the two ways of approaching language as contextual,
referring to facts outside language, and formal, referring to facts
inside language.
Contextual features are somewhere outside this physical realization of
the language. Streteches of language treated only formally are
referred to as text.
Now although it is true that we need to consider contextual factors we
cannot say that there are no formal links bet sentences in discourse.
We shall now try to categorize these formal links and then examine
how far they will go in helping to explain why a succession of
sentences is discourse, and not just a disconnected jumble.
Formal links bet sentences and bet clauses are known as Cohesive
Devices.

Verb forms: The form of the verb can limit the choice of the verb
form in the next.

Parallelism: Another link within discourse is effected by parallelism,

a device which suggests a connection, simpy because the form of one
sentence or clause repeats the form of another. This is often used in
speeches, prayers, poetry, and advertisements. It can have a
powerful emotional effect.
It doesn´t have to be necessarily grammatical parallelism. It may be a
sound parallelism; as in the rhyme, rhythm, and other sound effects
of verse. One might even extend the idea and talk of semantic
parallelism where two sentences are linked because they mean the
same thing.
Referring expressions: These are words whose meaning can only
be discovered by referring to other words or to elements of the
context which are clear to both sender and receiver. The most obvious
example of them is third person pronouns.
It is not only the third pers. pronouns which work in this way. The
meanings of this and that, and here and there have also to be found
either formally in another part of the discourse or contextually from
the world.

Anaphoric ref.: The identity of someone or something is given once at

the beginning, and thereafter referred to with pronouns.
Cataphoric ref.: The pronouns are given first and then the identity is
revealed.
Exophoric ref. : The meaning is foundcontextually from the outside
world.

Referring expresions fulfuil a dual purpose of unifying the text (they

depend upon some of the subject matter remaining the same) and of
economy, because they save us from having to repeat the identity of
what we are talking abt again and again.

Repetition and lexical chains: Repetition of wds can create the

same sort of chain as pronouns, and there are sometimes good
reasons for preferring it. In Britain, mother tongue learners of Eng are
discouraged from using repetition on the grounds that it is `bad style
´, and ecouraged to use a device known as `elegant repetition´,
where synonymous or more general wds or phrases are used. So
instead of writing:

The pineapple.. the pineapple.. the pineapple.. the pineapple

They might write

The pineapple.. the lusciuous fruit.. our meal.. the tropical luxury.

The kind of link that we choose will depend upon the kind of discourse
we are seeking to create, and elegant repetition is not always
desirable. It may sound pretentious in causal conversation, or create
dangerous ambiguity in a legal document.

We have described referring expressions, repetition, and elegant

repetition as establishing `chains´ of connected wds running through
discourse. Such lexical chains need not necessarily consist of wds
which mean the same, howver. They may also be created by wds
which associate with each other. This association may be by virtue of
some formal semantic connection (good, for example, associates with
its opposite bad), or it may be because wds are felt to belong to some
more vaguely defined lexical group (rock star; wld tour; millionaire;
yacht). This last kind of connection, though it is sometimes treated as
a kind of cohesion is really too dependent upon individual experience
and knowledge to be treated as a formal link.

Substitution: Another kind of formal link bet sentences is the

substitution of wds like do or so for a wd or group of wds which have
appeared in an earlier sentence: Do you like mangoes? Yes, I do. /
Yes, I think so.

Ellipsis: Sometimes we don´t even need to provide a substitute for a

word or phrase which as already been said. We can simply omit it,
and know that the missing part can be reconstructed quite
successfully. Instead of answering:
Would you like a glass of beer? With: Yes, I would like a glass of beer.
We can just say: Yes I would. The rest is understood.

Conjunction: Yet another type of formal relation bet sentences is

provided by those wds and phrases which explicitly draw attention to
the type of relationship which exists bet one sentence or clause and
another. These are conjunctions. They can simply add more info to
what has already been said (and, furthermore, add to that) or
elaborate or exemplify it (for instance, thus, in other wds). They may
contrast new info with old info, or put another side to the argument
(or, on the other hand, however, conversely). They may relate new
info to what has already been given in terms of causes (so,
consequently, because, for this reason) or in time (formerly, then, in
the end, next) or they may indicate a new departure or a summary
(by the way, well, to sum up, anyway).
They all indicate the relationship of utterances in the mind or in the
world and are thus in a way contextual.

3. Why formal links are not enough

Formal links bet sentences, then, are not enough to account for our
feeling that a stretch of lang is discourse. Tehy are neiher necessary
nor sufficient.

3.1 Language functions:

What kind of rules enable people to infer the function of what is said
from its literal, formal meaning?
In order to discover how such inferences are made, we will need firstly
to examine the range of possible functions of language, and secondly
to try to understand how people correctly interpret them.
Understanding this connection bet the form and the function of lang
will help us to explain how stretches of lang can be coherent without
being cohesive. We cannot assume that these interpretations will be
made in the same way in all cultures and in all languages, so
understanding how interpretation proceeds in the culture of the
language we are teaching is crucial if we are to help foreign learners
to make their wds function in the way that they intend.
From now on, we shall use the term utterance for a unit of language
used by somebody in context to do sth – to communicate – and
reserve sentence for grammatically complete units regarded purely
formally, in isolation from their context and their function.

3.2 The classification of Macro-functions:

There have been many attempts to classify the main functions of

language (macro-functions). One of the clearest and most influential
was formulated by the linguist Roman Jakobson (1960), and further
developed by Dell Hymes (1962).
The scheme proceeds by first identifying the elements of
communication [7] as follows:
The addresser: the person who originates the message. This is
usually the same as the person who is sending the message, but not
always, as in the case of messengers, spokepeople, etc.
The adressee: the person to whom the message is addressed. This
is usually the person who receives the message, but not necessarily
so, as in the case of intercepted letters, bugged telephone calls, etc.
The channel: the medium through which the message travels: sound
waves, marks on paper, telephone wires, etc.
The message form: the particular grammatical and lexical choices
of the message.
The topic: the information carried in the message.
The code: the language or dialect, for example, Swedish.
The setting: the social or physical context.

Macro-functions [7] are then established each focusing attention

upon one element:

The emotive function: Communicating the inner states and

emotions of the addresser (`Oh! No!´, `Fantastic!´)
The directive function: Seeking to affect the behaviour of the
addressee (`Please help me´, `Shut up!´).
The phatic function: Opening th channel or checking that it is
working, either for social reasons (`Hello, lovely day!´, `Do you come
here often?´) or for practical ones (`Can you hear me?´, `Are you still
there?´).
The poetic function: In which the particular form chosen is the
essence of the message (Beanz meanz Heinz).
The referential function: carrying information.
The metalinguistic function: Focusing attention upon the code
itself, to clarify it or renegotiate it (`What does this wd here means?´,
`This bone is known as the `femur´´).
The contextual function: Creating a particular kind of
communication (`right, let´s start the lecture´, `it´s just a game´).
3.3 Functional development:
It´s interesting to speculate, if one accepts this classification, on the
evolution of functions in each human individual. The crying baby is
being expressive, although her cries are not really language at all, but
instinctive reactions to the envirnometn. When she realizes that by
controlling these cries, and producing them at will rather than
automatically, she can influence the behaviour of her parents, she has
progressed to the directive function. Phatic communication also
begins very early. Chuckling, gurgling, babbling, often have no
function but to say: `Here I am, and so are you´. The poetic function is
also apparent at an early statge: when young children latch on to a
phrase and repeat it endlessly, without conveying any information.
The referential function gains its prominence only at a later stage,
and the metalinguistic function also comes later; these are the
functions on which a considerable amount of attention is lavished at
school.
Surprisingly, considering this course of development, a good deal of
foreign language teaching begins with the metalinguistic function, by
explicitly stating the rules of grammar.

3.4 Micro-functions and functional language teaching:

We might go on to subdivide each function and specify more delicate

categories, or micro-functions:.
A breakdown of the directive function for example:

Question
Requests for
action
Orders
Requests
for information
Directive function Requests
Requests
for help
Pleas
Requests
for sympathy
Prayers

Certainly no list could ever claim to be exhaustive and complete.

3.5 Functional analysis and coherence:

The idea of langugge function can go a long way towards solving this
problem of what binds utterances together as discourse in the
absence of formal links. If we can ascertain the function of utterances,
we will be able to perceive a unity of a different kind.
The important principle has been established, that meaning varies
with context. Formally, out of context, a sentence has a kind of time-
free and place-free meaning. Used as an utterance in context it may
have may meanings, which although they are connected to this
context-free sentence meaning, may be extremely varied.
These two types of meaning are distinguished by the terms semantic
meaning (the fixed context-free meaning) and pragmatic meaning
(the meaning which the wds take on in a particular context, bet
particular people. The function of an utterance must be established
pragmatically.
How do human beings interpret-usually quite accurately- what is
meant from what is said?And why does this divergence of function
and form exist at all? Why do people not just speak directly and say
what they mean? For an answer we4 will need to loos elsewhere: at
the theories of conversational principles and speech acts, ideas
which, as their names suggest, were developed with spoken language
in mind, but are as applicable to written discourse as to spoken.

3.6 Conversational principles: Co-operation:

The idea that conversation proceedss according to a principle, known

and applied by all human beings, was first proposed in a limited form
by the philosopher Paul Grice (1975), who put forward what he
described as the co-operative principle. According to this principle,
we interpret language on the assumption that its sender is obeying
four maxims. We assume he or she is intending to:

• Be True (the maxim of quality)

• Be Brief (the maxim of quantity)
• Be Relevant (the maxim of relevance)
• Be Clear (the maxim of manner)

Using this assumption, combined with general knowledge of the

world, the receiver can reason from the literal, semantic meaning of
what is said to the pragmatic meaning – and induce what the sender
is intending to do with his/her wds.

When we talk abt people following the co-operative principle, this

does not mean that they can consciously and explicitly formulate it to
themselves. People act as though they know the principle just as they
act as though they know the rules of grammar- though very few
people can even begin to formulate them, and nobody can formulate
them completely.
3.7 Flouting the co-operative principle:

There are times when meaning derives from deliberate violations – or

flautings as Grice calls them – of the co-operative principle, always
provided that the sender intends the receiver to perceive them as
such, and that this is how, in fact, the receiver does perceive them.
It is possible to flout the quality maxim without lying. For example:
`Queen Victoria is made of Iron´. Though it is not literally true, you
will perceive such remarks as figures of speech. These are Hyperbole,
metaphor, Irony and sarcasm and they will all depend upon the
assumption that they will be interpreted as deliberate floutings of the
charge to be true rather than as untruths intended to deceive. Just as
the quality maxim can be flouted for effect, so can the other three.
The quantity maxim is violated in both directions: creating prolixity if
we say too much and terseness if we are too brief. We often say more
than we need, perhaps to mark a sense of accasion, or respect; and
we often say less than we need, perhaps to be rude, or blunt, or
forthright.
Sometimes we deliberately flout the charge to be relevant, to sign
embarrasment or a desire to change the subject. Lastly, the maxim of
manner is violated either for humour, as in the case of puns and
doubles entendres, where rival meanings are deliberately tolearted,
or in order to establish solidarity bet speakers and exclude an
overhearer from the conversation.
The meanings created by these floutings are often social, signalling
the attitude of the sender to the receiver of the message, and the
kind of relationship which exists or is developing bet them. Grice
viewed these attitudinal meanings as being created by departures
from the co-operative principle.

3.8 Conversational principles: Politeness:

The politeness principle, like the co-operative principle, may be

formulated as a series of maxims which people assume are being
followed in the utterances of others. As with the co-operative principle
any flouting of these maxims will take on meaning, provided it is
perceived for what it is. Robin Lakoff formulated them as follows:

• Don´t impose
• Give options
• Make your receiver feel good.

These maxims of the politeness principle explain many of those

frequent utterances in which no new information is communicated. In
Eng we often give orders, and make requests and pleas (directives) in
the form of elaborate questions which give the option of refusal; we
apologize for imposing and add in praise to make our hearer feel
good. Clearly the politeness principle and the com-operative principle
are often in conflict with each other. Politeness and truth are often
mutually incompatible and so are politeness and brevity. These
conflicting demands of the two principles are something of which
people are consciously awarre. In english, there is even a term for the
surrender of truth to politeness: `A white lie´.

3.8 The social basis of conversational principles:

The co-operative and politeness principles, and the tension bet them,
reflect a dual purpose in human intercourse: to act efficiently together
with other people, and to create and maintain social relationships.
There are situations, and there are types of relationships, in which
one of these purposes becomes dominant, and the other hardly
matters at all. In emergencies, when there is a need for immediate
action, it is hardly appropriate to follow the politeness principle.
Brown and Levinson (1978), who have studied politeness phenomena,
suggest that teir origin is teh same in all societies. All human beings,
in order to enter into social relationships with each other, must
acknowledge the face of other people. By this they mean that people
both avoid intruding upon each other´s territory (physical territory, a
part field of knowledge, a friendship) and also seek to enlarge the
territory of others – in Lakoff´s terms, make the other person feel
good – presumably on the assumption that the same will be done to
them. The specific nature of face varies from society to society.
The precise way of indicating respect for face may be culture specific,
and not subject to direct translation. In some cultures, initial refusal of
an offer may be merely polite, and invite repetition; in others the
opposite may be true.
Thourgh their realizations differ, the two, often conflicting, aims of
communication – to co-operate and to maintain social relations – are
universal.
The theory also goes some wayy towards answering the question of
why people speak indirectly. It enables them to give options and also
to retreat behind the literal meaning of what is said.

3.9 Speech Acts:

Inferring the function of what is said by considering its form and

context is an ability which is essential for the creation and reception
of coherent discourse and thus for successful communication. The
principles of politeness and co-operation are not, on their own,
enough to provide the explanation for this inference. To do this we
also need knowledge of the physical and social world. We also need to
make assumptions abt the knowlege of the people with whom we are
interacting.
An approach which tries to formulate how such knowledge is brought
into play is speech act theory. This was first formulated by the
philosopher John Austin (1962). These ideas were further developed
by another philosopher, John Searle (1969, 1975), who both added to
them and presented them more systematically.

3.10 Declarations and performatives:

Speech act theory begins with the observation that there is a class of
highly ritualistic utterances which carryno information ab the wld
outside language at all, because they refer only to themselves.
Examples of such utterances are swearing an oath, sentencing a
criminal, opening a building, arresting a felon, naming a ship. They
are utterances in which saying the wds and doing the action are the
same thing: the function is created by the form. Such utterances are
labelled declarations.
However, the utterance only succeeds in having this function if
certain external conditions are fulfilled. The wds must be uttered by
someone with the necessary authority, in a country in which tere is a
death penalty, to a person who has been coniceted of a particular
crime; they must be spoken, not written, at the right time (at the end
of a trial) and in the right palce (in court). The conditions that must be
fulfilled are known as felicity conditions.

Declarations, however, are only a special case of a much commoner

group of utterances, performatives. These are also utterances in
which saying is doing, and they too are only successful if certain
felicity conditions are fulfilled, but, unlike declarations, their related
verbs (vow, arrest, declare, etc.) are not always actually said. A good
example is the act of ordering someone to do sth.

The felicity conditions of an order are:

1. the sender believes the action should be done.

2. the receiver has the ability to do the action.
3. the receiver has the obligation to do the action.
4. the sender has the right to tell the receiver to do the action.
(authority)

Conversely, we can see that if the conditions do hold, then any

reference by the sender to the action will be perceived as an order,
even without an explicit form like `I order you to ...´ or the imperative.

Speech act theory, which relates the function of utterances to sets of

felicity conditions and the knowledge of participants that these
conditions exist, may help us to understand the unity of exchanges in
communication.
3.11 Speech act theory and coherence:

Speech act theory provides us with a means of probing beneath the

surface of discourse and establishing the function of what is being
said.
We shall be able to examine the structure of discourse both in terms
of surface relations of form, and underlying relations of functions and
acts.

The family to which the act of ordering belongs is called directives.

Another family of speech acts which has been suggested is
expressives, which includes the act of thanking, apologizing,
welcoming, and congratulating. We also have the commissives, which
include acts of promising and threatening.

3.12 Underlying force:

Speech act theory uses technical terms for these layers of intention
and interpretation. The formal literal meaning of the wds is the
locution; the act which is performed by saying it the illocution; a third
layer is the perlocution or overall aim of the discourse. An utterance is
said to have illocutionary force and perlocutionary force. If we go back
to the private´s utterance `I´ve been scrubbing them all morning and
they won´t come any cleaner´, we can relate it to these three layers
as follows:

1. The locution: a statement conveying information that the

speaker has been cleaning his boots all morning.
2. The illocution: to challenge the sergeant´s order.
3. The perlocution: to undermine the sergeant´s authority, or to be
cheeky, or to escape the duty of cleaning the boots.

Notice how meaning becomes more and more slippery as we move

from one layer to the next. This is something which human beings
exploit to their advantage. It enables them to avoid commiting
themselves and to retreat in front of danger; and this is one of the
major reasons why people speak indirectly. Indirection also enables us
to give others the option of retreat. Quite often, people explicitly
discuss, or try to clarify the illocutionary and perlocutionary force, to
formulate the upshot of what is said.
Even in more casual situations people often try to get at the upshot of
what is being said with such utterances as `what are you trying to tell
me?´

3.13 Pragmatics, discourse analysis, and language teaching:

The fact that meaning is not constructed from the formal language of
the message alone is crucial in explaining what it is that makes
people perceive some stretches of language as coherent discourse
and others as disconnected jumbles. It is also important for the
successful teaching and learning of foreign languages.

The importance of pragmatic theories in lang learning is really

twofold. Firstly, the divergence of function and form means that we
cannot rely upon teachin only form. In production, learners need to
choose the wds which most suitable realize their intentiiioon, and this
does not always entail the most closely related form. There aer times
when making language function effectively is more important than
producing perfectly pronounced, grammatically correct sentences.
Secondly, the linking of form to function may help learners to
orientate themselves within a discourse.
The underlying structure of the discourse may be a progression of
functional units, and a breakdown in pragmatic interpretation may
easily add to a learner losing his/her way.
In order to `do things with wds´ either actively, as language
producers, or passively, as language understanders, we clearly need
more tools than the formal language system, though we do need that
too. The needs of the language user might be represented like this:

Pronunciation
Lang system Grammar
Vocabulary

Voice
USER Paralanguage Face GOAL
Body

Knowledge Cultural
World

Reasoning

Traditionally, language teaching has concentrated only on the hree

levels of the formal language system – pronunciation, grammar, and
vocabulary – and the way in which they function within the sentence,
on the assumption that other aspects of communication will follow
fairly automatically. It remains true, of course, that the formal system
needs to be acquired in some way. Is is not, however, all that is
needed for communication.
What we need to decide as language teachers is the degree to which
other components of communication need teaching. All human beings
have reasoning power, world knowledge, and knowledge of at least
one culture, but the divisions bet these categories, and the nature of
their contents are not always clear.

The pragmatic theories we have examined leave a number of

unasnwered questions. It is not always clear, for example, where the
context of an utterance ends and even when that is established we
are still left with the vagueness of the central concept of relevance.
Which elements of the context are relevant?. Anothe weakness is the
implicit assumption that underlying meaning can always be
formulated in wds. Speech act theory assumes that there is one neat,
verbally expressive illocution to each locution.

4. Two views of discourse structure as product and as

process

Pragmatics tends only to examine how meaning develops at a given

point. It provides us with something like a snapshot of meaning.
Discourse is more like a moving film, revealing itself in time –
sometimes over long periods.

4.1Rank structure

One way of representing the relationship of parts to a whole is as a

rank structure, in which each rank is made up of one or more of the
rank below. The ranks of grammar are
• Sentence
• Clause
• Phrase
• Word

Represented with the tree diagram

Sentence

Clause Clause

Phrase Phrase Phrase Phrase Phrase

Word Word Word Word Word Word

Word Word

Because such structures are conventional therefore culturally variable

the language learner, in order to be able to operate effectively as a
participant in discourse, needs to be able both to identify what type of
discourse he or she is involved in, and to predict how it will typically
be structured.

4.2The Birmingham School of Discourse Analysis

A pioneering and influential study in this field was carried out at the
University of Birmingham by Sinclair and Coulthard (1975). The
discourse type it chose to analyse was school lessons.

S & C recorded a number of British primary school lessons. On the

basis of these data they proposed a rank structure for these lessons
as follows:
• Lesson
• Transaction
• Exchange
• Move
• Act

They then drew up rules, based on the data, showing how these acts
combine together to form moves and how moves combine to form
various kinds of exchange – rather as grammarians formulate rules
describing how words combine into phrases, or phrases into clauses.
One kind of exchange, for example, consisted of between one and
three moves:
Opening (answering) (follow up)

4.3 Discourse typology: spoken and written; formal and informal:

Traditionally, language teaching has divided discourse into two major

categories, the spoken and the written, further divided into the four
skills of speaking and listening, writing and reading.

The traditional division of language into the spoken and the written is
clearly and sensibly based on a difference in production and
reception: we use our mouths and ears for one, and ours hands and
eyes for the other. Yet as far as discourse structure is concerned, a
more fundamental distinction seems to be bet formal, planned
discourse, which may be either written or spoken, and less formal,
unplanned discourse which – though it may also be either written or
spoken – is usually associated with speech. Informal spoken discourse
is something in which the modern foreign language learner, with
opportunities for travel and social contact, is most likely to wish to
succeed, but also the discourse type he or she is likely to find hardest,
precisely because it is so informal and unpredictable.

4.4Conversation as a discourse type:

The term `Conversation´ is widely used, in a non-technical sense,

usually with the implication that the talk is less formal. We shall
define the term as follows: Talk may be classed as conversation when:
It is not primarily necessitated by a practical task.
Any unequal power of participants is partially suspended.
The number of participants is small.
Turns are short.
Talk is primarily for the participants and not for an outside audience.

Although these definitions are imprecise, they are useful.

The boundary bet conversation and other discourse types is a fuzzy
one, and there are many intermediate cases. A seminar, for example,
might come somewhere bet the two poles.

Formal spoken discourse------------------------------------------------

Conversation

4.5Turn – Taking:

Overlap of turns occurs in only abt 5 per cent of conversation or less,

strongly suggesting that speakers somehow know exactly when and
where to enter. Where there is overlap bet turns it has some particular
signnificance: signalling annoyance, urgency, or a desire to correct
what is being said. Conversely, pauses bet turns also carry particular
meaning.
The significance of this approach for the langugage learner is
considerable. Turn-taking mechanisms, the way in which speakers
hold or pass the floor, vary bet cultures and bet languages. Overlap in
a given situation is more or less tolerated in some societies than in
others.
Efficient turn-taking also involves factors which are not linguistic. Eye
contact is one strong means of signalling, and in British culture (in
very general terms) it can often be observed that speakers look away
during their turn and then look their interlocutor in the eye at the end.
Body position and movement also play and important part. Intonation
and volume contribute to turn-taking too.
The relative status of the speakers, or the role which one of them is
playing, are also important. In formal situations roles can clearly give
people special rights, but even in conversation – where according to
our definition unequal power is suspended – it is unlikely that
knowledge of participants´ social status will be wholly forgotten.
Students fall silent when the professor speaks – in the bar as well as
in the seminar.

4.6Turn types:

One kind of turn alternation the ethnomethodologists describe is an

adjacency pair. This occurs when the utterance of one speaker makes
a particular kind of response very likely. A greeting, for example, is
likely to be answered by another greeting. In an adjacency pair, there
is often a choice of two likely responses. A request is most likely to be
followed by either an acceptance or a refusal. In such cases, one of
the responses is termed the preferred response (because it occurs
most frequently) and the other the dispreferred response (because it
is less common).

1. Offer Acceptance (preferred)

Refusal (dispreferred)

2. Assessment Agreement (preferred)

Disagreement (dispreferred)

3. Blame Denial (pr.)

Admission (dispr.)

4. Question Expected answer (pr.)

Unexpected ans. (dispr.)

Sometimes the second part of an adjacency pair can be delayed by an

alternation of turns occurring within it.

e.g.

A: Did you enjoy the meal?

B: (Did you?
A: yes)
B: So did I.

This is known as an insertion sequence.

The topic of an insertion sequence is intimately related to that of the
main sequence in which it occurs. Sometimes, however, speakers
simply switch from one topic to another unrelated one, and then back
again. In this latter case the insertion is known as a side sequence.

Insertion and sequences draw attention to the fact that conversation

is discourse mutually constructed and negotiated in time. Unlike
lectures, broadcasts, and speeches, a conversation is constructed and
executed as it happens, by two people, feeling their way forward
together. There is no going back, crossing out, rewriting and
restructuring. This is particularly evident in the phenomenon known
as repair, in which participants correct either their own words or those
of another participant, edging towards a situation in which maximum
communication is achieved.

4.7 Discourse as a process:

Ethnomethodology depicts conversation as discourse constructed and

negotiated bet the participants, following pre-established patterns,
and marking the direction they are taking in particular ways: with
pauses, laughter, intonations, filler words, and established formulae.
Culture specific rules and procedures of turn-taking provide ample
breeding ground for misunderstanding. Entering and leaving
conversation, bidding for a longer turn, refusing without appearing
rude, changing the topic, are all notoriously difficult for foreign
learners: tasks for which the language classroom, where turns are
patiently organised and controlled by the teacher, has hardly
prepared them. Indeed the teacher who constantly itnerrupts the
students´ discourse to correct every grammatical mistake not only
violates usual turn-taking procedures but may also hinder the
students´ acquisition of them.

5. Discourse as dialogue.

5.1 Dialogue in communnicative development:

Developmentally, dialogue comes first, both for the human specied,

and for the human individual. We have no hard evidence of the origins
of language in prehistoric communities, but is seems reasonable to
assume that speech preceded writing and dialogue preceded
monologue.

As with societies, so with each individual infant. Turn-taking and

interaction are among the first communicative skills.

5.2Disourse typology: reciprocity:

In discourse analysis it has been fairly common to distinguish bet two

fundamental types of discourse: reciprocal and non-reciprocal.
Discourse is reciprocal when there is at least a potential for
interaction, when the sender can monitor reception and adjust to it-
or, to put it in another way, where the receiver can influence the
development of what is being said. In non-reciprocal discourse,
sender and receiver may have no opportunity for interaction. The
prototype of non-reciprocal discourse is a book by a dead author. The
distinction, however, is misleading.
Absolutely non-reciprocal discourse is unlikely. Even writers working in
solitude try to form some idea of the receiver of their work and adjust
to it- the meaningfulness of what they say can be viewed as a
measure of the success of that predicition and adjustment.

Reciprocity is a question of degree. All discourse is more or less

reciprocal, if only because it is based upon assumptions abt receivers.
It should also be clear that although there is a general tendency for
speech to be more reciprocal and writing to be less so, this is by no
means necessarily true, and the reciprocal-non-reciprocal cline, like
the formal-informal cline, cuts across the distinctionbet speech and
writing. A monarch´s speech at a state opening of parliament though
spoken, is farm from the reciprocal end of the scale, but a scribbled
memo from one teacher to another, though written, may trigger off a
series of replies and counter replies, and is thus highly reciprocal.

5.3Information structure in discourse:

One widely accepted explanation is that the ordering of information is

determined by the sender´s hypotheses abt what the receiver does
and does not know. With interpretation we might divide information
into two types – that which the sender thinks the receiver already
knows, and that which the sender thinks the receiver does not already
know – and label these two types given information and new
information respectively. Any unit of information may of course
change status as the discourse proceeds, and what was new in one
sentence becomes given in the next, precisely because it has just
been said. Indeed, communication might be defined as the conversion
of new information into given information, and a successful
communicator as a person who correctly assesses the state of
knowledge of his or her interlocutor. If we misjudge, and treat what is
given as new, we will be boring; in the reverse case when we assume
the new to be given, we will be incomprehensible.

A typical discourse, then, proceeds roughly as follows:

Given....... New. Given...... New. Given....... New.

Each given unit being already known by the receiver, or deriving from
a preceding piece of new information.

Our choices among the options for arranging the information are
neither arbitrary, nor just aesthetic devices to ensure variety, but
have some communicative funciton, making discourse more readily
comprehensible.
As we do make important choicies bet alternative versions of
sentences, even though each one is correct in itself, then in a
succession of sentences, it is possible that the choice is being
dictated by the sentence before, each one having a knock-on effect
on the structure of the next. At first then, it would seem that this
ordering of information is another instance of a formal connection bet
sentences in discourse. On closer inspection it turns out to be also
contextual, dictated by what is going on in the mind of the sender and
the assumptions he or she makes about what is going on in the mind
of the receiver.
One way of understanding this is to view the discourse as proceeding
by answering imagined and unspoken questions by the receiver. In
this light, all discourse seems to proceed like a dialogue, even if the
other voice is only present as a ghost.
 6. Knowledge in discourse

The choices we make abt the order of the information in discourse

reveal our own assumptions abt the world and abt the people we are
trying to communicate with. The truth of those assumptions gives
unity to our discourse and success to our communication. Their
falsehood puts it in danger of collapse.

6.1 Evidence for Schemata:

There are a number of pieces of evidence that the mind does in fact
employ knowledge schemata in the interpretation of discourse.
One piece of evidence is the fact that people questioned abt a text or
asked to recall it, frequently fill in details which they were not actually
given, but which a schema has provided for them.

6.2Complex schemata:

Not surprisingly, considering the complexity of the interaction of

minds, language, and the world, the description we have given so far
is highly simplified. Acutal discourse is unlikely to be interpretable
with reference to a single schema. In reality the mind must activate
many schemata at once, each interacting with the other. It must be
capable of moving rapidly from one to another, of using more than
one simultaneously, of focusing on a sub-schema (say a menu-
schema within a `restaurant-schema´). It must be capable of builiding
new schemata, and of ditching old ones.
Participants in conversation have certain –- no doubt highly culture-
bound – assumptions abt possible courses for a conversation, length
and type of turn, total duration, and so on. Less reciprocal discourse
will also activate schemata.

6.3 Relevance:

Schemata, then, are data structures, representing stereotypical

patterns, which we retrieve from memory and employ in our
understanding of discourse. The successful communicator selects just
those features which differ from this schema, enabling the receiver to
adjust it and to bring it closeer to the individual instance which is
being described.
Schema theory can go a long way towards explaining the sender´s
choice and arrangement of information in communication. One of
Grice´s maxims tells us to `be relevant´- but it doesn´t attempt to
explain the notion of relevance.
Sperber and Wilson (1986) have used a model of communication
which is very closely related to schema theory to explain the concept
of relevance. Human minds, they say, have a long-term aim: to
increase their knowledge of the world. In each encounter with
discourse, we start with a set of assumptions, whose accuracy we
seek to improve. Information is relevant when it allow us to alter our
knowledge structures to give us a more accurate representation of
the world. On the other hand, successful communication must work
within the framework of the receiver´s exisiting knowledge; it must
not make too many demands. Successful communication gives us
new information, but works within the framework of the receiver´s
assumptions.

Schema allow human communication to be economical. It would be

hard to see how communication could take place if we could not take
some sort of mutually shared knowledge for granted, if every
discourse had to begin from scratch.
Communication also suffers when people make false assumptions abt
shared schemata.
Misjudgements and mismatches of schemata are particularly likely
when people try to communicate across cultuers and across
languages. For this reason schema theory is of as great importance in
language teaching as it is in discourse analysis.

6.4 Discourse deviation:

Trying to understand the process by which two or more people come

together through text to create discourse and thus communicate can
be a very stimulatin and exciting investigation. But there are also
times when it can seem depressing. Increasingly, we seem to be
talking abt the unity and meaningfulness of discourse in terms of
conformity: to another human being, it seems that the most
successful communication will take place where there is already a
considerable coincidence bet mental states, and teh alteration
achieved is only minimal. People who see the world differently, and
therefore need to communicate, both for mutual education and to
avoid conflict, may seem the least likely to be able to do so.

What happens to those who step outside the predictable patterns and
regularitites? Strangely, some are vilified and some are glorified.
Some are called mad, disturbed, maladjusted, rebellious, even
criminal; others are called individualists, poets, comedians,
philosophers. It is easy to escape this issue by saying that the
discourses of the two groups have little in common; but discourse
analysis should teach us that it is as likely to be our attitude to what
they say that categorises them. Yet however we may judge deviation,
whether negatively or positively, being a social outsider is very much
a case of non-conformity to the norms and regularities of discourse
structure.

The discourse strategies of a foreign speaker may seem refreshing

exactly because they do not conform to conventions of the culture
whose language they are learning; on the other hand they may cause
serious misunderstanding and breakdown of communication. The task
of the language teacher is a difficult one: to facilitate a degree of
socialisation which will enable learners to send and receive text as
discourse, while also guarding their right to be different and to enrich
others through that difference, bringing to the language they are
learning the wealth of their own individuality and culture. As in the
case of deviation within the social group: we do not have to judge
difference negatively.

Success in communication depends as much upon the receiver as on

the sender. Between speakers of different languages it depends as
much upon the native speaker as on the foreign learner.