DISCOURSE

ANALYSIS
THREE WAYS OF LOOKING AT
DISCOURSE
Three different perspectives based on three different definitions of what discourse is.
Some have taken a formal approach to discourse, defining it simply as ‘language
above the level of the clause or sentence.’ Those working from this definition often try
to understand the kinds of rules and conventions that govern the ways we join clauses
and sentences together to make texts.
Others take a more functional approach, defining discourse as ‘language in use’. This
definition leads to questions about how people use language to do things like make
requests, issue warnings, and apologize in different kinds of situations and how we
interpret what other people are trying to do when they speak or write.
Finally, there are those who take what we might call a social approach, defining
discourse as a kind of social practice. What is meant by this is that the way we use
language is tied up with the way we construct different social identities and
relationships and participate in different kinds of groups and institutions. It is tied up
with issues of what we believe to be right and wrong, who has power over whom, and
what we have to do and say to ‘fit in’ to our societies in different ways.
Although these three different approaches to discourse are often treated as separate,
and are certainly associated with different historical traditions and different
individual discourse analysts, they are better seen as three interrelated aspects of
discourse. The way people use language cannot really be separated from the way it is
put together, and the way people use language to show who they are and what they
believe cannot be separated from the things people are using language to do in
particular situations.

CONTEXT
Perhaps the most famous model of context is that developed by the linguist Michael
Halliday, who drawing heavily on the work of both Malinowski and Firth, also
proposed a three‐part model of context.
For him, context consists of:
1) Field: the social action that is taking place;
2) Tenor: the participants, their roles and relationships;
3) Mode: the symbolic or rhetorical channel and the role which language plays in the
situation.
Halliday goes a bit further than Firth in explaining the relationship between context
and actual language use with his concept of register.
By register, Halliday means the different ways language is used in different situations
in terms of things like the content of what is said and the degree of formality with
which it is said. The basic distinction of register is between spoken language and
written language.

CONTEXT AND COMPETENCE

Halliday explains context from an essentially linguistic point of view, seeing it as part
of a language’s system of ‘meaning potential’.
‘There is no need to bring in the question of what the speaker knows,’ he writes; ‘the
background to what he does is what he could do – a potential, which is objective,
not a competence, which is subjective. In sharp contrast to his position is that of the
linguistic anthropologist Dell Hymes, for whom the notion of ‘competence’ is central
to a model of context he called ‘the ethnography of speaking’, or, as it is sometimes
called, ‘the ethnography of communication’.
In his work, Hymes focused on the interaction between language and social life ‐ the
ways using and understanding language are related to wider social and cultural
knowledge. Knowledge or mastery of the linguistic system alone, he insisted, is not
sufficient for successful communication. People also need to know and master various
rules, norms and conventions regarding what to say to whom, when, where, and
how-
which he called communicative competence.
He wrote:
The sharing of grammatical (variety) rules is not sufficient. There may be persons
whose English I can grammatically identify but whose messages escape me. I may
be ignorant of what counts as a coherent sequence, request, statement requiring an
answer, requisite or forbidden topic, marking of emphasis or irony, normal duration
of silence, normal level of voice, etc., and have no metacommunitative means or
opportunity for discovering such things. (Hymes 1974: 49)

COhesion and cohErence

One of the most basic tasks for a discourse analyst is to figure out
what gives text and conversations texture. Texture, we said, comes from
cohesion and coherence. Cohesion primarily has to do with linguistic features
in the text, and coherence has to do with the kind of ‘framework’ with which the
reader approaches
the text and what he or she wants to use the text to do.

COHESION
Halliday and Hasan describe two broad kinds of linguistic devices that are used to
force readers to engage in this process of backward and forward looking which
gives them a sense of connectedness in texts. One type depends on grammar (which
they call grammatical cohesion) and the other type depends more on the
meanings
of words (which they call lexical cohesion)
Devices used to create grammatical cohesion include:
・Conjunction (using ‘connecting words’)
・Reference (using a pronoun to refer to another word)
・Substitution (substituting one word or phrase for another word or phrase)
・Ellipses (leaving something out)

Lexical cohesion involves the repetition of words or of words from the same
semantic field (e.g. milk, tomatoes, rocket).

Conjunction refers to the use of various ‘connecting words’ (such as conjunctions

like and and but and conjunctive adverbs like furthermore and however) to join
together clauses and sentences. Conjunction causes the reader to look back to the first
clause
in a pair of joined clauses to make sense of the second clause. The important thing
about these ‘connecting words’ is that they do not just establish a relationship
between the two clauses, but that they tell us what kind of relationship it is.
‘Connecting words’, then, can be grouped into different kinds depending on the
relationship they establish between the clauses or sentences that they join together.
Some are called additive, because they add information to the previous clause or
sentence. Examples are ‘and’, ‘moreover’, ‘furthermore’, ‘in addition’, ‘as well’. Others
are called contrastive because they set up some kind of contrast with the previous
sentence or clause. Examples are ‘but’, ‘however’. Still others are called causative
because they set up some kind of cause and effect relationship between the two
sentences or clauses. Examples of these are ‘because’, ‘consequently’, ‘therefore’.
Finally, some are called sequential because they indicate the order facts or events
come in. Examples are ‘firstly’, ‘subsequently’, ‘then’ and ‘finally’.
All connecting words cause the reader to look back to a previous clause (or sentence)
in order to understand the subsequent clause (or sentence), and the kind of
connecting word used guides the reader in understanding the relationship between
two clauses (or sentences).
Another very common way we make our texts ‘stick together’ is by using words that
refer to words we used elsewhere in the text. This kind of cohesive device is known as
reference. What reference does is to help the reader to keep track of the various
participants in the text as he or she reads (Eggins, 1994: 95).
There are basically three kinds of reference:

1) anaphoric reference – using words that point back to a word used before:

After Lady Gaga appeared at the MTV Music Video Awards in a dress
made completely of
meat , she was criticized by animal rights groups.
 2) cataphoric reference: Using words that point forward to a word that has not
been used yet:

When she was challenged by reporters, Lady Gaga insisted that the
dress was not intended to offend anyone.

3) Using words that point to something outside the text (exophoric reference):

If you want to know more about this controversy, you can read the
comments people have left on animal rights blogs.

The definite article (the) can also be a form of anaphoric reference in that it usually
refers the reader back to an earlier mention of a particular noun.

Lady Gaga appeared in a dress made completely of meat. The dress

was designed by Franc Fernandez.

Substitution is similar to reference except rather than using pronouns, other

words are used to refer to an antecedent, which has either appeared earlier or will
appear
later. In the sentence below, for example, the word one is used to substitute for
dress.

Besides wearing a meat dress, Lady Gaga has also worn a hair one,
which was designed
by Chris March.

Substitution can also be used to refer to the verb or the entire predicate of a clause,
as in the example below.

If Lady Gaga was intending to shock people, she succeeded in doing so.

Ellipsis is the omission of a noun, verb, or phrase on the assumption that it is

understood from the linguistic context. In order to fill in the gap(s), readers
need to look back to previous clauses or sentences, as in the example below.
There is much to support the view that it is clothes that wear us, and
not we, them. (Virginia Woolf)

All of the devices mentioned above are examples of grammatical cohesion, the
kind of cohesion that is created because of the grammatical relationship between
words.
Lexical cohesion occurs as a result of the semantic relationship between words.
The simplest kind of lexical cohesion is when words are repeated.
But a more common kind is the repetition of words related to the same subject. We
call these ‘chains’ of similar kinds of words that run through texts lexical chains.
In the following text, for example, besides the use of reference (who, it, she), the
clauses are held together by the repetition of the verb ‘to wear’ and of other words
having to do with clothing and fashion (bikini, Vogue – a famous fashion magazine,
dress, and outfits).

Lady Gaga, who came under fire recently for wearing a meat bikini on
the cover of Vogue Hommes Japan, wore a raw meat dress at last
night's VMAs. It was one of many
outfits she wore throughout the night. (Oldenberg, 2010)

Taken together, these words form a lexical chain, which helps to bind the text
together.
Lexical chains not only make a text more cohesive but also highlight the topic
or topics (such as ‘fashion’, ‘entertainment’, ‘technology’) that the text is about – and
so can provide context for determining the meaning of ambiguous words.

COHERENCE

What makes a text, a text is often as much a matter of the interpretative framework
that the reader brings to the text as it is of anything internal to the text.
The relationship between the words ‘tomatoes’ and ‘rocket’ becomes meaningful to a
reader based on his or her understanding of what a shopping list is and what it is used
for.
This aspect of texture is known as coherence, and it has to do with our
expectations about the way elements in a text ought to be organized and the kinds of
social actions (like shopping) that are associated with a given text. There are a
number of different kinds of interpretative frameworks that we use to make sense of
texts. A generic
framework is based on the expectations we have about different kinds of texts, the
kinds of information we expect to encounter in texts of different kinds and the order
in which we expect that information to be presented, along with other kinds of lexical
or grammatical features we expect to encounter.
Part of what forms such generic frameworks is that different parts of a text are not
just grammatically and lexically related, but that they are conceptually and
procedurally related – in other words, that they appear in a certain logical or
predictable sequence.
But not all of the knowledge we use to make sense of texts comes from our knowledge
about the conventions associated with different kinds of texts. Some of this knowledge
is part of larger conceptual frameworks that we build up based on our understanding
of how the world works. We will use the term cultural models to describe these
frameworks. James Paul Gee (2010) calls cultural models ‘videotapes in the mind’
based on experiences we have had and depicting what we take to be prototypical
(or ‘normal’) people, objects and events.
The important thing to remember about cultural models (and, for that matter,
generic frameworks) is that they are cultural. In other words, they reflect the
beliefs and values of a particular group of people in a particular place at a particular
point in history.