Discourse Analysis

Reading list

Discourse analysis
• Brown, Gillian and George Yule.1983. Discourse Analysis. Cambridge:
CUP.
• Chafe, Wallace, 1994. Discourse, Consciousness, and Time. Chicago: The
University of Chicago Press.
• Johnstone, Barbara. 2008. Discourse Analysis. Malden,MA/London:
Blackwell Publishing.
• Liddicoat, Anthony J. 2007. An Introduction to Conversation Analysis.
London/ New York: Continuum.
• Schiffrin, Deborah. 1994. Approaches to Discourse. Cambridge, MA:
Blackwell Publishers Inc.
• Schiffrin, D., Tannen, D. and Hamilton, H.E. (eds.). 2001. The Handbook of
Discourse Analysis. Malden, MA and Oxford: Blackwell.
• Tsui, A.M.B. 1994. English Conversation. Oxford: OUP.
• van Dijk, T. (ed.). 1997. Discourse as Structure and Process. London and
Thousand Oaks, CA: Sage Publications, Inc.
• van Dijk, T. (ed.). 1997. Discourse as Social Interaction. London and
Thousand Oaks, CA: Sage Publications, Inc.
• Woods, Nicola. 2006. Describing Discourse: A Practical Guide to
Discourse Analysis. London: Hodder Arnold.
Corpus linguistics
• Kennedy, Graeme. 1998. An Introduction to Corpus Linguistics.
London: Addison Wesley Longman Ltd.
• McEnery, Tony et al. 2006. Corpus-Based Language Studies: An advanced
resource book. London & New York: Routledge.
• Meyer, Charles F. 2002. English Corpus Linguistics: an introduction.
Cambridge: CUP.
• Teubert, Wolfgang & Krishnamurthy (eds.), R. 2007. Corpus Linguistics (6
volume set). London & New York: Routledge.
Research methodology
• Wray, Alison & Bloomer, Aileen. 2006. Projects in Linguistics:
A Practical Guide to Researching Language. London: Hodder
Arnold.
Suggested projects

• Pragmatic strategies: how to make suggestions; how to give

advice; how to make requests; how to make apologies; how to
make/pay compliments; how to persuade people; politeness;
impoliteness; innuendo
• Language phenomena: repetition; repair; overlap;
interruption; reported speech; return pop
• paralinguistic phenomena: laughter; silence
1. Book review
1) Choose a book published after 2000
2) Write a critical review of the book
2. Survey article
Choose one of the following and write a survey of it:
media discourse, classroom discourse, academic discourse, children’s
discourse, gender discourse , forensic discourse, workplace discourse, discourse
grammar, discourse comprehension, discourse structure, discourse markers,
genre, reference, cohesion, coherence , etc.
3 ． Research article
1 ） Choose a research topic in discourse analysis
2) Collect your own data (either spoken or written)
What is discourse?
discourse
talk (spoken discourse) text (written discourse)

discourse analysis text linguistics

Discourse studies are about talk and text in

context.
What is discourse analysis? (DA)
• Michael Stubbs(1983)
The term discourse analysis is very ambiguous. I will use it in this book to refer
mainly to the linguistic analysis of naturally occurring connected speech or
written discourse. Roughly speaking, it refers to attempts to study the
organization of language above the sentence or above the clause, and
therefore to study larger linguistic units, such as conversational exchanges or
written texts. It follows that discourse analysis is also concerned with
language use in social contexts, and in particular with interaction or
dialogue between speakers (Stubbs:1).
In his book, discourse analysis is defined as:
1) concerned with language use beyond the boundaries of a
sentence/utterance,
2) concerned with the interrelationships between language and society and
3)concerned with the interactive or dialogic properties of everyday
communication.
What is discourse analysis? (DA)
• Teun A. van Dijk

Three dimensions of discourse analysis

1. language use
2. the communication of beliefs (cognition)
3. interaction in social situations
• When we focus on discourse as a form of language
use, that is, on the verbal aspect of utterances,
linguistics usually makes a metaphorical distinction
between different levels of such utterances. Once we
know a bit more about discourse as language use, we
move to a characterization of its communicative and
interactional dimensions.
• Discourse as verbal structure
1)Sound, sight and body: two different modes of discourse (talk
and text)
2) Order and form: discourse syntax; discourse and grammar
3) Meaning: proposition, the micro level of analysis, the macro

level of analysis
4) Style: stylistic analysis
5) Rhetoric: rhetorical analysis
6) Schemata: schematic structures or superstructures
• Discourse as Action and Interaction
Speech acts
Conversation as interaction
• Cognition
Mind, memory
Mental processes and representations
• Discourse and society
Social power
Gender
Ethnicity
Culture
CDA
Types of discourse analysis
1) those which focus on discourse “itself”, that is , on structures
of text or talk;
2) those which study discourse and communication as
cognition
3) those which focus on social structure and culture

This triangle of discourse—cognition—society is indeed the

site of multidisciplinary discourse analysis
The Emergence of DA

• Zellig Harris 1952 Discourse Analysis

• In the mid 1960s

The Interdisciplinary Nature of DA
• From anthropology, discourse analysis takes a concern with the
embedding of language and language use in a wider sociocultural
context. Anthropologists study the diversity of human cultures, often
using the method of participant observation (observing a community
while participating as much as possible in its activities oneself) to
produce a kind of description that is known as ethnography. A
researcher undertaking ethnographic work in any community must try
to understand that community’s culture----its ways of acting in the
world and making sense of the world---in the way community
members understand it themselves. Speaking is both a way of acting
in the world and a means for making sense of it, and language has
thus been one of the aspects of culture that anthropologists have paid
attention to.
The Interdisciplinary Nature of DA
• From philosophy, discourse analysis takes a concern with the
way language acquires meaning when it is used. DA has drawn
in particular on the tradition known as “ordinary language”
philosophy (the philosophical study of ordinary language as
opposed to symbolic “languages” such as formal logic).
Ordinary language philosophers such as J. L. Austin, John
Searle and Paul Grice drew attention to a feature that seems to
be characteristic of human linguistic communication, that we
can “mean more than we say” (or less than we say, or something
different from what we say). To interpret utterances in
discourse, we have to be able to do more than just decode the
meaning of the words: we have to work out how the speaker
intends us to take the utterance.
The Interdisciplinary Nature of DA
The approach to discourse analysis that has developed from
the tradition of ordinary language philosophy is known as
pragmatics. Pragmatics concerns itself with the principles
language-users employ to determine the meaning behind
words---how we get from what is said to what is meant.
The Interdisciplinary Nature of DA
• From sociology, discourse analysis takes a concern to
account for the orderliness of social interaction. The question
of how social order in general is produced and reproduced is
traditionally a central concern of sociology. One particular
approach to this general sociological question---the approach
known as ethnomethodology, and associated in particular with
the theorist Harold Garfinkel---gave rise to the way of
analyzing talk that is now known as CA. CA is, among other
things, an enquiry into the nature of the procedures
conversationalists follow to produce the orderliness of
ordinary talk.
The Interdisciplinary Nature of DA
• From linguistics, DA takes a concern with the structure of
language and the distribution of linguistic forms. Interest in the
formal and structural properties of spoken interaction remains
prominent in some traditions of DA. For example, in the 1970s a
group of linguists studying classroom interaction elaborated a
model of “exchange structure” ,pointing out that when teachers ask
questions in the classroom, the resulting exchanges typically have a
three-part structure which the analysts termed “elicitation-
response-feedback”. The teacher asks a question, a pupil produces
a response, and the teacher takes another turn in which he/she
makes clear whether the pupil’s response is acceptable (Sinclair
and Coulthard 1975). This contrasts with what happens in ordinary
conversation, where asking a question typically initiates a two-part
exchange, “question-answer”.
The Interdisciplinary Nature of DA
A rather different example of a structural approach to discourse is found
in the work of the sociolinguist William Labov, best known for his
pioneering work on language variation and change. Labov has also made
important contributions to discourse analysis, proposing an influential
model of the structure of spoken narrative. Labov asserted that the
fundamental problem of discourse analysis is to show how one utterance
follows another in a rational, rule-governed manner.
Interactional sociolinguistics is another kind of structural approach
with roots in linguistics, because it is concerned with the distribution of
particular features in talk (where in talk you find them, and whose talk
you find which ones in). At the same time the approach has roots in
anthropology too: people’s different ways of doing things interactionally
may be analyzed in relation to their cultural beliefs, assumptions and
values.
The Interdisciplinary Nature of DA
• There is also an approach to discourse analysis, known as
Critical Discourse Analysis, which borrows its conceptual
and analytic apparatus from both structural linguistics and
intellectual enterprise sometimes known as critical theory.
CDA is an approach that focuses on how reality (including
such aspects of it as power and gender relations) is
constructed in and through discourse by analyzing actual
examples closely, and importantly, by paying attention not
only to their content but also to their form.
some basic principles of doing discourse analysis
1. Naturally occurring text and talk
2. Contexts
3. Discourse as talk
4. Discourse as social practice of members
5. Member’s categories
6. Sequentiality
7. Constructivity
8. Levels and dimensions
9. Meaning and function
10. Rules
11. Strategies
12. Social cognition
Approaches to DA
• 1.The Ethnography of Speaking
• 2.Conversation Analysis
• 3.Pragmatics
• 4.Interactional Sociolinguistics
• 5.Critical Discourse Analysis
• 6. Variation Analysis
• 7. Systemic Functional Grammar
• 8. Relevance Theory
• ……
1. Situations and events: the ethnography of speaking( 言语人
类文化学 ) /the ethnography of communication

What is the ethnography of speaking?

The term ethnography belongs to the specialist vocabulary of
anthropology 人类学 ; it refers to the investigation of culture(s) using a
particular methodology, that of participant observation. Typically,
anthropologists go and live with the people they wish to study for an
extended period of time. They learn about the community’s way of life
both through observational techniques (e.g. recording things that happen,
interviewing people )and through participating as much as possible in
community activities themselves. They are simultaneously “inside” the
culture, immersed in its day-to-day life, and “outside” it , trying to
understand the way its members think and act, and reflecting on their
own progress towards that goal.
the ethnography of speaking
Participant observation is also used by researchers who are not
anthropologists, but who are working in fields like sociology, cultural
studies or sociolinguistics. For example, a sociologist studying youth
subcultures might decide that instead of just interviewing young
people individually or in groups about their subcultural practices, they
should seek out one or more groups prepared to let a researcher spend
time with them regularly: join in their activities, record their
conversation, and talk to them about what they do and why. This
resembles the methods used in anthropological fieldwork, and is
therefore labeled “ethnography”.
Language is a relevant consideration in all ethnographic research,
since participant observation inevitably involves, among other things,
talking to the people you are studying and observing their talk among
themselves, which is to say, it produces at least some data in the form
of spoken discourse.
the ethnography of speaking
Linguists are familiar with the distinction, introduced by N. Chomsky in the
1960s, between linguistic competence and performance.
In the 1970s the anthropologist and linguist Dell Hymes pointed out that a
“competent” language user needs to know more than just a set of rules for
forming grammatical sentences; s/he also needs to know how to use language in
a contextually appropriate way. It is not by chance that people regularly produce
the right kind of utterance at the right time to the right person. Hymes proposed
an analogue for Chomsky’s “linguistic competence”, which he called
communicative competence. Linguistic competence is about rules of
grammar; communicative competence is about rules of speaking. The
ethnography of speaking developed as a way of investigating the rules of
speaking that are operative in particular language-using communities.
If one is investigating rules of grammar, the objects of analysis will be
grammatical units such as sentence, clause, phrase. What, though, are the
relevant units for an investigation of rules of speaking? Hymes proposed three,
and like sentences, clauses and phrases in grammar they are hierarchically
ordered.
the ethnography of speaking
The highest-level unit is the speech situation, the social
context in which speaking takes place. Speech situations
provide occasions for the use of language, but they are not
purely linguistic. For example, a family meal might be
analyzed as a speech situation , but other activities feature in
it besides talking (e.g. eating , drinking, serving food and
drink, feeding infants, kicking people under the table ,etc.)
and the overall proceedings are not governed only by rules of
speaking.
the ethnography of speaking
The next unit down from the speech situation is the speech
event. Speech events are constituted by the use of language:
they involve activities which could not occur except in and
through language, such as ‘argument’, ‘gossip’, ‘storytelling’.
Typically, community members have metalinguistic labels for
recurring speech events; they say things like ‘I went over to
Jeannie’s for a chat’, ‘today’s lecture was really boring’, ‘the
bus was late and Mrs. Smith gave me a good telling off, ‘how
did your interview go?’, It will be apparent that more than one
speech event might occur within the same speech situation:
‘chat’, ‘argument’ and ‘telling-off’, for instance, are
commonly observed in the course of a single family meal.
the ethnography of speaking
The ‘lowest’ level unit of analysis in this framework is the speech act.
Speech acts are things like ‘greeting’, ‘apologizing’, ‘insulting’,
‘asking/answering a question’. They are not speech events in
themselves, and in many cases (e.g. ‘greeting’) they can figure in many
different kinds of events; but the distinctiveness of a particular speech
event is partly a question of which acts are performed and in what order.
While all three levels are analytically relevant, the most important
is the midlevel one, the speech event, since it is essentially the event to
which ‘rules of speaking’ apply. Hymes proposed that speech events have
a set of ‘components’, characteristics which the analyst needs to look at
in order to produce a satisfactory description of any particular speech
event. In the work of Hymes and others we find various lists in which the
components are named and ordered in slightly different ways, but the
listing most often used to introduce this descriptive framework is a
mnemonic device that labels each component with one of the letters of
the word speaking.
the ethnography of speaking
• S setting: where the speech event is located in time and space.
• P participants: who takes part in the speech event, and in what role( e.g. speaker,
addressee, audience, eavesdropper 偷听者；窃听者 )
• E ends: what the purpose of the speech event is, and what its outcome is meant
to be.
• A act sequence: what speech acts make up the speech event, and what order they
are performed in.
• K key: the tone or manner of performance (serious or joking, sincere or ironic,
etc.)
• I instrumentalities: what channel or medium of communication is used (e.g.
speaking, writing, whistling etc.) and what language/variety is selected from the
participants’ repertoire.
• N norms of interaction: what the rules are for producing and interpreting speech
acts.
• G genre: what ‘type’ does a speech event belong to, and what other pre-existing
conventional forms of speech are drawn on or ‘cited’ in producing appropriate
contributions to talk (e.g. do people quote from mythology or poetry or
scripture?)
the ethnography of speaking
This set of components is sometimes referred to as the
‘SPEAKING grid’, and its purpose is to help analysts put
their observations in some kind of order. Someone observing
a culture to which s/he is an outsider is likely, at least initially,
to observe various things– including unfamiliar speech events
– whose significance s/he does not fully understand. In this
situation it can be useful to have some kind of preexisting
descriptive framework to put your observations in. Without
such a framework you may miss important things, or interpret
things in terms of categories that are used by your own society
rather than the one you are observing. Hymes’s grid is meant
to be both comprehensive and applicable to any community’s
ways of speaking.
 2. Sequence and structure: Conversation Analysis

In spite of its name, Conversation Analysis does not

deal only with conversation: the approach has also been
applied to talk in professional and workplace settings, to
political speeches and to media genres such as radio
phone-in programs. Many practitioners label their object
of study not “conversation” but “talk-in-interaction”.
Conversation Analysis
The fact that talking is prototypically a joint
enterprise involving more than one person, and
that people normally take turns at talk, is central
to the CA approach, which is concerned above
all to describe sequential patterns (regularities in
what follows what) which are observable in the
data being analyzed. Hence, CA is a markedly
“data centered” form of discourse analysis.
Conversation Analysis
In trying to explain how it is that
conversation can happen at all, Sacks et al.
found it necessary to account for two “grossly
apparent facts ”that they observed about
spoken interactive data. These facts are that:
a.only one person speaks at a time.
b.speaker change recurs.
Conversation Analysis
These two facts lie behind our commonsense
observation of conversation, that it is, fundamentally, a
turn-taking activity. In trying to explain how it is that
speakers keep taking turns, the Conversation Analysts
modeled conversation as a generative mechanism,
designed to fulfill two distinct functions. First,
speakers have to be able to work out when it is
appropriate to transfer the role of speaker. Second,
there has to be a way of determining who the next
speaker is to be.
Conversation Analysis
Sacks et al. suggested that speakers recognize
points of potential speaker change because
speakers talk in units which they called Turn
Constructional Units (TCUs). They defined a
TCU as a grammatically complete unit of
language, such as a sentence, clause or phrase,
the end of which represents to the interactants a
point at which it is possible for speaker transfer
to occur.
Conversation Analysis
The second problem of who should be the next
speaker also depends on TCUs. Sacks et al. note that at
the end of a TCU there are two possibilities for
determining the allocation of turns. The first
possibility is that the current speaker selects the person
who is to be the next speaker. Typical ways by which
the current speaker can select the next speaker include
the use of names or vocatives, gaze, posture and the
targeting of moves such as directing questions to
particular interactants. The second possibility is that
the next speaker may self-select. The system of turn
allocation is summarized as follows:
Conversation Analysis
• The system of turn allocation
• select a different
• speaker
• current speaker
• choose next speaker
• select self

• Point of possible
turn-transfer

next speaker
self-select
Conversation Analysis
Sacks et al. point out that this system operates at the end
of every turn or what they call “locally”, rather than on an
overall or “global” basis. In other words, turn allocation
cannot be agreed in advance at the beginning of the
conversation, but must be continually renegotiated at each
TCU boundary. The system has one aim: to ensure that
when the current speaker finishes his turn at talk, some
other speaker will start talking. The conversation analysts
thus modeled conversation as an infinitely generative turn-
taking machine, whose design suggests that the major
concern of interactants is to avoid lapse: the possibility that
no one is speaking.
Conversation Analysis
Given that conversation is driven by a turn-taking
machinery expressly designed to keep going, it
becomes quite problematic to determine how a
conversation could ever stop. It was in exploring an
answer to this issue of conversational closure that CA
made what many see as their most significant
contribution to the analysis of interaction: the
identification of the adjacency pair.
Conversation Analysis
Sacks et al. noticed that the occurrence of the
second turn can be explained by the first: a question in
some ways implies that the next turn will be an answer.
Adjacency pairs were identified as typically having
three characteristics:
a. two utterance length
b. adjacent positioning of component utterances
c. different speakers producing each utterance
Conversation Analysis
Adjacency pairs were identified as functioning as a turn-
transfer technique, i.e. they function both to allocate the
next turn, and to exit from the current turn. It is important to
understand that the system is not one of determination,
but of expectation. Given that a first pair part has been
produced, there is a very strong likelihood that the
addressed participant will be the next speaker, and will
produce a relevant second pair part.
e.g.
A: Who was the first president of the United States?
B: George Washington.
Conversation Analysis
The conversation analysts recognized that in fact there
are two types of second pair parts. There is first a
preferred second pair part. e.g. question/answer,
demand/compliance, offer/accept, request/grant etc.
However, it is also possible for the addressee to
produce some kind of discretionary alternative--- what
in CA terms is referred to as a dispreferred second pair
part.
Conversation Analysis
e.g.
(1) A: Who was the first president of the United States?
B: You could ask my five-year-old kid.

(2) A: Are you still having some sort of contact with your
former husband?
B: I’m afraid that’s a rather personal question.
Conversation Analysis
• The identification of the adjacency pair
was the basis for two further developments
in CA. The first was the recognition of
sequences longer than two units, and the
second was the formulation of the theoretical
concept of sequential relevance/sequential
implicativeness.
Conversation Analysis
• e.g.
1. C: And have you got…the… first day covers 首日封 of…?
2. S: Yes.
3. C: (Anzac)
4. S: How many would you like?
5. C: Four please.
6. S: Two of each?
7. C: What have you got?
8. S: There’s two different designs on the-
[5 secs---S shows C the covers]
9. C: I’ll take two of each.
10.S: uhum.
Conversation Analysis

Here we see that turns 1 & 2 constitute a

question/answer adjacency pair. However, this
pair is followed by a sequence of seven turns
which in some sense seems to “go together” or
have a particular relatedness to each other. It takes
these seven utterances for the salesperson’s
question (How many would you like?) to be
resolved so that the customer’s initial request can
be complied with.
Conversation Analysis
Instances such as these prompted CA to recognize
the existence of sequences, of which the adjacency pair
was merely the minimal version. In their description of
sequences, CA has frequently concentrated on those
which are particularly “visible” in that they interrupt,
suspend, or prepare for the ongoing interaction. The
types of sequences referred to most frequently in the
CA literature are: insertion sequences(1&2),
pre-sequences(3 &4), side sequences(5), etc.
Conversation Analysis
• e.g.
• (1)
A: May I have a bottle of Mich? (Q1)
B: Are you twenty-one?(Q2)
A: No. (A2)
B: No. (A1)
(2)
A: I don’t know where the---this address is. ( Q1)
B: Well, where do---which part of town do you live? (Q2)
A: I live four ten East Lowden. (A2)
B: Well, you don’t live very far from me. (A1)
(3)
A: Do you have blackberry jam?
B: Yes.
A: Okay. Can I have half a pint then?
B: Sure.
Conversation Analysis
(4)
A: Did you hear the terrible news?
B: No. What?
A: You know Grandpa Bill’s brother Dan. He was killed
in a car accident this morning.
(5)
A: If Percy goes with Nixon, I would sure like that.
B: Who? (misapprehension)
A: Percy. That young fellow. His daughter was murdered. (clarification)
B: Oh, yeah, yeah. (terminator)
Conversation Analysis

The side sequence is different from the

insertion sequence. While the insertion
sequence is made up of a pair, the side sequence
is made up of more than one adjacency pair. A
side sequence contains three parts:
misapprehension, clarification and terminator.
Conversation Analysis
From the analysis of adjacency pairs and turn
sequences came a recognition of the more general
principle underlying conversational organization: that
of the sequential relevance or sequential
implicativeness of talk. This is the notion that
conversational turns make sense because they are
interpreted in sequence. As Atkinson and
Heritage(1984:6) point out: ‘no empirically occurring
utterance ever occurs outside, or external to, some
specific sequence. Whatever is said will be said in
some sequential context’.
Conversation Analysis
CA suggests that the most significant placement
consideration in conversation is that of adjacency. Thus
whenever possible the speaker’s current turn will be
interpreted as implicating some action by the responder
in the immediate next turn. Similarly, the respondent’s
subsequent talk will, where possible, be interpreted as
related to the immediately prior turn. Thus, adjacency
pairs can be seen as merely the prototypical variety
of the general conversational principle of sequential
relevance. This principle explains why we work
extremely hard to interpret any two adjacent turns as
related, despite the absence of any other indications of
cohesion.
Conversation Analysis
Turn-taking organization and adjacency pair
organization are two kinds of local organization. Now
let’s turn to the overall organization of conversation.

When conversation analysts study the overall

organization of conversation, they mainly focus on the
opening and closing structures. Openings tend to
contain a greeting, an enquiry after health and a past
reference (as in “how did it go last night?”)
Conversation Analysis
e.g.
Brenda: Hi,Lee.
Lee: Hi. Hi, Jean.
Jean: Hi,hi.
Brenda: How are you?
Lee: Not bad. I’ll be in, in a minute.
Conversation Analysis

Schegloff & Sacks(1973) point out that

closings of conversation contain three
fundamental components: closing sequence
(The interactants exchange such farewell
expressions as “Goodbye, Good night, See you”
etc.), pre-closing sequence (It is indicated by
the use of such expressions as “Alright, Okay,
So, Well” etc.) and topic bounding sequence.
Conversation Analysis
e.g.
A: Yeah, well. Things always work out for the best.
tbs B: Certainly.
A: Alright, Tess.
pcs B: Uh huh. Okay.
A: Goodbye.
cs B: Good night.
One kind of conversation with a recognizable overall
organization that has been much studied is the telephone call.
Opening of a phone call

Phone rings in Beth’s location

Beth: hello
Ann: hello
Beth: oh hello Ann what’s up?
Ann: nothing much just had something I wanted to ask you.
 Closing of a phone call

• Here is an extract from a call made by Jane , who is trying

to get five people ,including Paula, together for a meeting
on a Saturday. At this point Jane has elicited the
information she needs, regarding which Saturdays in the
next few weeks Paula is free.
• Transcription conventions: (.)brief pause,? rising intonation,:
lengthened segment
1.Jane: so that’s any Saturday except the seventh(.) right I’ll let
you know when I’ve heard from others
2.Paula: ok
3.Jane: right then
4.Paula: I might be in London before that though shall I give you a
ring?
5.Jane: yeah do that
6.Paula: ok I will.
7.Jane: good
8.Paula: ok::
9.Jane: take care then
10.Paula: you too
11.Jane: bye
12.Paula: ok bye
• At turn 3 in the sequence above, Jane has an opportunity to go
on to some matter other than the business the pair has just
transacted, but what she says is merely “right then”. This is
potentially a “pre-closing”, an indication that she is ready to
close and an invitation to Paula to pass on her own next turn.
Paula, however, does not pass. She brings up the previously
unmentioned topic of her intention to visit London. Jane is
thus obliged to respond to the new topic, and Paula completes
the sequence by agreeing to ring when she makes the visit. At
turn 7 Jane again indicates readiness to close by saying
“good”. This time Paula also passes: she says “OK”, with
marked lengthening on the last syllable which carries the
implication of finality. With the pre-closing successfully
accomplished by the two women each passing up the
opportunity to introduce new matter, the actual closing can
begin. It takes them ten turns to get from the end of their main
business (at turn 2) to the end of the call.
This extract gives us the impression that the end of the call is
“dragging on”. It is not uncommon for the end of a call to be far
more extended than it is in this example.
 Institutional talk/workplace interaction
Business meetings, service encounters, doctor-patient consultations,
talk in courtrooms or classrooms etc.
Drew and Heritage(1992:22)summarize the main differences between
institutional and ordinary talk as follows:
1) Institutional interaction involves an orientation by at least one of
the participants to some core goal, task or activity (set of them)
conventionally associated with the institution in question. In short,
institutional talk is normally informed by goal orientation of a
relatively restricted conventional form.
2) Institutional interaction may often involve special and particular
constraints on what one or both of the participants will treat as
allowable contributions to the business at hand.
3) Institutional talk may be associated with inferential frameworks
and procedures that are particular to specific institutional
contexts.
Transcription conventions: italics=emphasis, (6)pause duration in
Seconds, - = short pause, [ =simultaneous speech
Magistrate: I’m putting it to you again - are you going to make an
offer – uh – uh- to discharge this debt(6)
Defendant: Would you in my position
Magistrate: I – I’m not here to answer questions – you answer my
question
Defendant: One rule for one and one for another – I presume
Magistrate: Can I have an answer to my question – please(6)
The question is - are you prepared to make an offer to the court – to
discharge- this debt.
Defendant: What sort of minimal offer[ would be required
Magistrate: [it’s not a bargaining situation – it’s a
straight question Mr. H.- can I have the answer
3. Doing things with words: Pragmatics
Speaking as doing
Two philosophers, John Austin and John Searle,
developed speech act theory from the basic belief that
language is used to perform actions: thus, its
fundamental insights focus on how meaning and action
are related to language. This school of philosophy is
called ‘ordinary language philosophy’, which
flourished in England in the middle of the 20th century.
The proponents of this school wanted to analyze
philosophical problems by looking at ordinary
language and trying to ascertain what insights it could
offer into reality.
 Although speech act theory was not first
developed as a means of analyzing discourse,
some of its basic insights have been used by
many scholars to help solve problems basic to
DA. In addition, particular issues in speech act
theory lead to DA, e.g. how an utterance can
perform more than one speech act at a time, and
the relationship between context and
illocutionary force.
Pragmatics
There is a striking difference between the
following two sentences:
e.g.
(1)It’s raining.
(2)I promise that I will give you one hundred
dollars tomorrow.
 In (1) a statement is made that may or may
not be true. As for (2), however, it is not
possible to say that it is true or that it is not true.
With verbs such as ‘promise’ (in the first
person), something is not only being said; more
importantly, something is being done. In (2) an
act is being performed in the form of an
utterance. By saying ‘I promise…’, a promise is
made. But saying ‘It’s raining’ does not make it
rain.
Pragmatics
Austin used the terms ‘constative ’ and
‘performative’ to describe this difference. In
constatives, such as (1), something is said about reality;
in performatives, such as (2), an act is performed by
the utterance itself. Austin was not ,however,
successful in establishing criteria for describing the
difference between these two concepts. It can, after all,
be argued that an act is being performed in the case
of constative utterances as well; an indirect request( a
request to the hearer to hand the speaker her umbrella)
made or a statement (factual assertion about the world)
made as in the case of (1).
 In recognition of the fact that utterances can both make
propositions and perform actions, Austin came to the
conclusion that all expressions of language must be viewed
as acts. He distinguished three kinds of action within each
utterance. First, there is the ‘locution’, the physical act of
producing an utterance. Second, there is the ‘illocution’ , the
act which is committed by producing an utterance: by uttering
a threat, a threat is made; by uttering an apology, an apology
is made. Third, there is the ‘perlocution’, the production of an
effect through locution and illocution, for instance, the
execution of an order by the hearer.
Pragmatics
In speech act theory, the illocution is the
focus of attention. Certain minimum
requirements must be met if an illocution is to
be successful. The American philosopher John
Searle formulated four ‘felicity conditions’
which illocutions must meet. These four
conditions will be illustrated using the illocution
‘to promise’.
• Felicity conditions for the speech act ‘to promise’
a. propositional content condition
S predicates some future act A of S.
b. preparatory condition
S believes that A is in H’s best interest and that S can
do A.
c. sincerity condition
S intends to do A.
d. essential condition
• S puts S under an obligation to H to do A.
Pragmatics
Searle used felicity condition to show that
the successful exchange of speech acts is also
bound by certain rules. In terms of form and
function, this means that a form can only
acquire a valid function given certain
conditions.
 How does speech act theory contribute to the study of
discourse? First, it can provide insights into the requirements
which the production of a form (the locution) must meet to
ensure that the illocution takes place. This illocution serves as
a prerequisite for the achievement of the intended perlocution.
Second, this theory can serve as a framework for indicating
what is required in order to determine the relationship
between form and function, between locution and illocution.
Below is an example in the form of an interrogative.
e.g.
(3) Can you stop by in a minute ?
Pragmatics
Why is this interrogative generally interpreted as a
request? A request can be identified by the following
felicity conditions:
Felicity conditions for requests
a. propositional content condition
The content must refer to a future act, x, which is
to be carried out by the hearer.
b. preparatory condition
1. The hearer is capable of executing x and the
speaker believes that the hearer is capable of
doing it.
2. It is obvious to both conversational participants
that the hearer will not perform the act without
being asked.
c. sincerity condition
The speaker actually wants the hearer to do what
has been requested.
d. essential condition
The utterance serves as an attempt to persuade the
hearer to execute x.
Pragmatics
On the basis of rules in this definition, it can be
said that the interrogative given in (3) possesses the
illocutionary intent of a request. This does not,
however, explain why this interrogative must be
interpreted as an order when it is uttered by a
supervisor to a subordinate. In this case, the situation is
not self-explanatory and a knowledge of the
surrounding environment is required. Compare the
following examples.
e.g.
(4) This panther has brownish-yellow spots.
(5) Your left eye has brownish-yellow spots.
Both cases can be viewed as simple statements, but
(5) can also be intended as a warning if a situation is
being described which can be viewed as dangerous. It
could, on the other hand, also be seen as a sign of
affection. Again, knowledge of the world is required to
be able to deduce this.
Pragmatics
• There are a number of cases in which the utterance
itself provides an indication of the intended illocutions.
Searle calls these indications IFIDs, illocutionary
force indicating devices. IFIDs include performative
verbs, word order, intonation, accent, certain adverbs,
etc. If an IFID is present, the utterance is said to have
an explicit illocution; in all other cases the utterance is
said to have an implicit or indirect illocution. Below
are a few examples of explicit illocutions.
(6) I request that you put out your cigarette.
(7) He is putting out his cigarette.
(8) Is he putting out his cigarette?
(9) Are you going to put that cigarette out or
not?
Pragmatics
In (6) the performative verb ‘request’ makes
the illocutionary intent explicit. The difference in
word order between (7) and (8) is indicative of
the illocutionary intent, in this case ‘statement’
and ‘question’ respectively. Ascending intonation
and an accent on the word ‘cigarette’ can also
convey an expression of surprise. In (9) the tag
“or not” is indicative of the imperative character
of the illocution.
 It should be noted that IFIDs do not always provide
a definitive answer regarding illocutionary intent. The
IFID “if… then” in the following two examples would
suggest a conditional promise, but in fact there is a
conditional promise only in (10).
(10) If you take the garbage out, I will give you a
beer.
(11) If you keep this up, you will have a nervous
breakdown.
Pragmatics
• In (11) the IFID is not the only relevant
factor; more background information is needed,
specifically that a nervous breakdown is
dangerous. Otherwise, it is impossible to deduce
why (11) is generally seen as a warning. If so
much additional information is needed to
determine the function of explicit language
utterances, then it should be clear that this is
even more difficult in the case of implicit or
indirect utterances.
Pragmatics :The cooperative principle
The Danish linguistic philosopher Otto
Jespersen wrote in the introduction to his
Philosophy of Grammar(1924):
The essence of language is human
activity---activity on the part of one individual
to make himself understood by another, and
activity on the part of that other to understand
what was in the mind of the first.
 If two parties use an instrument for an activity, then such
an activity can only be successful if both parties adhere to
general rules or principles and thereby utilize certain
strategies. This can be illustrated with a non-linguistic
example. If two people want to hang a painting (activity),
they use a hammer, nails, and a ladder (instruments), and they
have to coordinate their actions. There will have to be some
form of cooperation; while one is standing on the ladder, the
other can hand the tools to the first, etc. Rules concerning
politeness will also have to be followed; while one person is
on the ladder, the other should not try to push the first off.
One general principle of collective activity is ‘cooperation’,
and an often used strategy to achieve this is ‘politeness’. This
is also true in the case of verbal communication.
Pragmatics: The cooperative principle
A speaker’s words often convey more than the literal
meaning of the words uttered. The following example
is from the classic article Logic and Conversation
(1975) by the logician and philosopher Paul Grice.
(12) Suppose that A and B are talking about a
mutual friend, C, who is now working in a bank. A
asks B how C is getting on in his job, and B replies, Oh
quite well, I think he likes his colleagues and he
hasn’t been to prison yet.
 The form of this utterance does not say everything
about the meaning and therefore the function. A can
derive from B’s remark that B does not hold a high
opinion of C. In fact, B has basically said that C is a
potential criminal. Yet, this cannot be derived from the
literal meaning of B’s words. Why then can A draw
these conclusions? Because A can assume that there is
some relevance to B’s, at first glance, superfluous
addition concerning prison. The only reason B would
add that remark is if B meant to imply that C is a
potential criminal.
Pragmatics
Grice called this derivation ‘conversational
implicature’. By using the term ‘implicature’,
Grice wanted to emphasize that it is not a
logical implication such as the if-then
relationship. The addition of the word
‘conversational’ denotes that the derivations
being dealt with are an essential part of the
information-transfer process in conversations.
 A speaker can only get a meaning like this across if the
listener cooperates. To capture this notion, Grice
formulated a general principle of language use, ‘the
cooperative principle’:

The cooperative principle

Make your conversational contribution such as is
required, at the stage at which it occurs, by the accepted
purpose or direction of the speech exchange in which you
are engaged.
Pragmatics
CP consists of four specific maxims.
Quantity
1) Make your contribution as informative as is required
(for the current purposes of the exchange ).
2) Do not make your contribution more informative than is
required.
Quality
Try to make your contribution one that is true.
1) Do not say what you believe to be false.
2) Do not say that for which you lack adequate evidence.
Relation
Be relevant.
Manner
Be perspicuous
1) Avoid obscurity of expression.
2) Avoid ambiguity.
3) Be brief (avoid unnecessary prolixity)
4) Be orderly.
 The maxims of the cooperative principle can be
used to describe how participants in a conversation
derive implicatures. Grice gives the following example.
A is standing by an obviously immobilized car and is
approached by B. The following exchange takes place.
(13) A: I am out of petrol.
B: There is a garage round the corner.
A can deduce from B’s reaction that B means that
there is a garage around the corner that is open and
sells gasoline. B, however ， has not mentioned these
facts. A can only make these assumptions if s/he
assumes that B is acting in accordance with the
cooperative principle and is adhering to the maxim of
relevance.
• The important point made by Grice regarding the theory of
implicature is the distinction between what is said and what is
implicated.

• Conversational implicature is based on the Cooperative Principle.

• In Grice’s analysis, the speaker’s flouting of a maxim combined

with the hearer’s assumption that the speaker has not really
abandoned the co-operative principle leads to an implicature.
flouting the maxim of quantity

A writes a reference for X who applies for a

philosophy job.
Dear Sir,
Mr. X’s command of English is excellent
and his attendance at tutorials has been
regular.

Yours …
• A: Where is Bill?
• B: There is a yellow VW outside Sue’s house.
flouting the maxim of quality
Everybody knows A betrays B.
• A is a fine friend. (irony)
• Queen Victoria was made of iron.
(metaphor)
Someone broke all the furniture.
• He is a little intoxicated. (meiosis)
• Every nice girl loves a sailor. (hyperbole)
flouting the maxim of relevance
• A: The hostess is an awful bore,
don’t you think?
B: The roses are lovely, aren’t they?
flouting the maxim of manner

A: Let’s get the kids something.

B: Okay, but I veto I-C-E-C-R-E-A-M.
Generally, conversational implicatures can be
divided into two kinds: generalized and
particularized.
Generalized conversational implicatures
No special knowledge is required in the context to
calculate the additional conveyed meaning.
A: I hoped you brought the bread and the cheese.
B: Ah, I brought the bread.
A: b & c
B: b +> NOT c

• A: Did you invite Mary and Cathy? (b & c?)

B: I invited Mary. (b +> NOT c)
Particularized implicatures
• Particularized implicatures are generated in a
conversation by overtly and deliberately
violating some submaxims for communicative
purposes.
features of conversational implicatures

cancellability
Dear Sir,
Mr. X’s command of English is excellent and his
attendance at tutorials has been regular. However, his
ability at and enthusiasm for
philosophy are quite adequate for the job.

Yours …
 non-detachability
The conversational implicature is attached to the semantic content of
what is said, not to the linguistic form used.
Therefore it is possible to use a synonym and keep the
implicature intact. In other words, the implicature will not be
detached, separated from the utterance as a whole,
even though the specific words may be changed.
e.g.
a-d said ironically will all implicate e
a. John is a genius.
b. John is a mental prodigy.
c. John is an enormous intellect.
d. John is a big brain.
e. John is an idiot.
• Calculability

With the literal meaning or the sense of

the utterance on the one hand, and the
cooperative principle and the maxims
on the other, it follows that an addressee would
make the inference in question to preserve the
assumption of cooperation.
• non-conventionality

A conversational implicature is the adding

up of the conventional meaning and the
context in which it is used. When the context changes, the
implicature will change or be lost.
e.g. It is cold here.
The sentence may, in one context, generate
an implicature like “Close the window”, but, in another
context, if it is said between two
small boys in the open, there will be no such
implicature. So the implicature is not
conventional.
• Indeterminacy
This refers to the fact an expression with a single
meaning can produce different
implicatures as the contexts and
participants vary.
e.g. He is an ox.
If the sentence is said by his superior in praising
him, it will imply that he is tough and hard
working, while in another
context it may mean that he is mild, but
stubborn: “stubborn as an ox” or stupid: “dumb ox”
 In discourse studies the cooperative
principle and its maxims are often referred to as
they provide a lucid description of how listeners
(and readers) can distill information from an
utterance even though that information has not
been mentioned outright. This is of importance
to research on the relationship between form
and function.
Pragmatics
Grice did, however, have a number of
additional comments concerning the cooperative
principle. First, the maxims are only valid for
language use that is meant to be informative.
This excludes, for example, such categories as
debating and small talk. Second, there are, from
the esthetic or social point of view, other
possible maxims. Grice suggests the maxim ‘be
polite’.
 Third, another principle is at work here.
Consider the quantity maxim. An
overabundance of information does not
necessarily have to mean that it is this maxim
that is being violated, since it can also be seen
as a waste of time and energy and thus as a
violation of some efficiency principle. Fourth,
some maxims are rather vague. For example,
how can it be determined which information is
required (first maxim of quantity)?
Pragmatics
• In the literature on Grice’s maxims special attention
is given to the maxim of relevance. It is unclear how it
can be determined whether a contribution to a
conversation is relevant or not. A number of
suggestions have been made in the direction of a clear
description of relevance. It has, however, proved to be
exceedingly difficult to determine exactly when the
maxim of relevance has been violated. Take the
following example of a question and a number of
possible answers.
• (14) A: Where’s my box of chocolates?
B: (a) Where are the snows of yesteryear?
(b) I was feeling hungry.
(c) I’ve got a train to catch.
(d) Where’s your diet sheet?
(e) The children were in your room this morning.
Pragmatics
A could react with surprise and ask why B is
suddenly quoting a line of poetry, in the case of
answer (a), or with I was talking about
chocolates and now you’re talking about the
children , in the case of answer (e). At first
sight, it seems that B is not acting within the
constraints of the maxim of relevance.
However, if A assumes that B is adhering to the
maxim of relevance, then any reaction B gives
could be construed as being relevant
 (a) B is not just quoting poetry; B is not really asking a
question. B, by reacting the way he does, is simply making
clear that the chocolates, like the snows of the past, have
gradually disappeared and that there is no good answer to
A’s question
(b) B is making clear that he has eaten A’s chocolates
(c) B does not want to answer the question because he is in a
hurry. Or, B is evading the question with an excuse; he
knows more than he is letting on .
.
(d) B is postponing giving an answer; first he wants to know
whether or not A should be eating chocolate.
(e) B is suggesting that the children ate the chocolates. Or, B is
suggesting that the children know where the chocolates are.

Obviously numerous other possible reactions for B are

conceivable. The main point is that every reaction can be
construed as being relevant
Pragmatics: politeness strategies
The cooperative principle is valid for informative language
use. Language users are not however, always interested in the
effective transfer of information. In the following examples
the speaker wants the hearer to close the door.
(15a) Close the door.
(15b) There’s a draft.
(15c) Would you close the door?
(15d) Would you be so kind as to close the door?
According to the maxims of the cooperative principle, (15a)
is sufficient. Language is, however, often used more
indirectly, as is done in (15b). They also sometimes use
certain politeness forms such as in (15c) and (15d).
 An important source of inspiration in the study of
politeness phenomena is the work done by Erving Goffman.
This social psychologist introduced the concept of ‘face’ .By
this he meant the image that a person projects in his social
contacts with others. Face has the meaning as in the saying
‘to lose face’. In Goffman’s opinion , every participant in the
social process has the need to be appreciated by others and the
need to be free and not interfered with. Goffman calls the
need to be appreciated ‘positive face’ and the need not to be
disturbed ‘negative face’.
Pragmatics
• Goffman wanted social interaction, which includes verbal
communication, to be studied from the perspective that
participants are striving for stability in their relationships with
others. Participants in conversations should, therefore, not
violate one another’s face. Refusing a request or reproaching
someone are actions which can form a threat to the other’s
positive or negative face. In the case of ‘face threatening acts’,
something is needed which will reduce the violation of face to
a minimum and therefore preserve stability as much as
possible. This can be achieved by using ‘face work
techniques’.
 How does politeness fit into this approach?
Politeness prevents or repairs the damage caused by
FTAs. The greater the threat to stability, the more
politeness, face work technique, is necessary. Just as
there are two types of face, there are two types of
politeness. Face work that is aimed at positive face is
called ‘solidarity politeness’, while face work that
deals with negative face is known as ‘respect
politeness’.
 Below are a few examples. When a personnel
manager has to turn down a job applicant who should
not have applied in the first place owing to a lack of
education, this is an FTA which threatens the positive
face of the applicant. For this reason the personnel
manager will be more apt to write (16b) than (16a).
(16a) We do not understand why you bothered to apply.
(16b) We have some doubts concerning your prior
education.
 In the following interaction between an instructor and a
student at the end of a tutoring session, the second variant is
more polite as it is less damaging to the instructor’s face and
that of the student.
(17) A: I’ve tried to explain this as clearly as possible .Now I
have to leave as I have another appointment. I hope
that the homework will be easier next time.
B: (a) I still don’t understand the material.
• (b) If problems should arise, is it all right if I stop
by
• tomorrow.
Pragmatics
Inspired by Goffman’s work, Penelope Brown and Stephen Levinson
(1978) developed a theory on the relationship between the intensity of the
threat to face and linguistically realized politeness. The intensity of the
threat to face is expressed by a weight (W) that is linked to an FTA. This
weight is the sum of three social parameters: a) the rate of imposition,
which is the ‘absolute weight’ of a particular act in a specific culture; b)
the social distance between the speaker and the person addressed; c) the
power that the person being spoken to has over the speaker. The term
‘absolute weight’ refers to the fact that, for example, the request “ May I
borrow your car ?” is in a category other than “May I borrow your pen?”
The request to borrow a car is of course not quite such a heavy demand if
the person requesting the car is the car owner’s brother. This illustrates
that the factors ‘distance’ and ‘power’ influence the ultimate weight.
Pragmatics
The ultimate weight of a FTA can be expressed by a value according
to the formula:
• Intensity of threat to face
W(FTA) =R ＋ D ＋ P
Brown and Levinson do not indicate how values are to be assigned to
R (rate of imposition), D (distance), and P (power). But it should be
clear that the value for P is different in the following examples.
(17a) Excuse me, sir, would it be all right if I smoke?
(17b) Mind if I smoke?
Utterance (17a) is more likely to be said by an employee to his boss,
while in the same situation,(17b) might be said by the boss to the
employee. In these examples the parameters R and D have the same
values.
Pragmatics
• In their research on linguistically realized politeness, Brown and Levinson
have investigated a number of languages. Their analyses indicate that there
are many ways of committing an FTA with a given weight. All of these
variants can, according to Brown and Levinson, be reduced to five strategies:
• possible strategies for doing FTAs
• 1.without redressive action, baldly
• on record 2. positive politeness
Do the FTA with redressive action
• 3. negative politeness

• 4. off record

5. Don’t do the FTA

Pragmatics
The fifth strategy is implemented when the risk of speaking is too great,
when, for instance, an individual does not risk answering an impertinent
and face-threatening question and simply remains silent.
Below are examples of different strategies for asking a person for a
hundred dollars.
(18a) Hey, lend me a hundred dollars. (baldly)
(18b) Hey, friend, could you lend me a hundred bucks? (positive politeness)

(18c) I’m sorry I have to ask, but could you lend me a hundred dollars?
(negative politeness)
(18d) Oh no, I’m out of cash! I forgot to go to the bank today. (off record)
The strategies are numbered according to their degree of politeness.
(Strategy 5 is, from this point on, left out of consideration.) If the W of a
FTA is high, the speaker will choose a strategy with a higher number.
This explains why grave accusations or inconvenient requests are often
formulated indirectly (strategy 4).
Pragmatics
The Goffman approach and Brown and Levinson’s theory provide an adequate
research framework for determining gradations of politeness and for analyzing
indirect language. The following question is an example of an indirect request.
(19) Are you doing anything special tonight?
The form of this utterance makes it clear that this is an inquiry about an
individual’s planned activities. This question can, however, also be an invitation
on the part of the speaker to the hearer to go out together.
How can a question in this form have an entirely different function? According
to Levinson (1983), the answer is that in some cases speakers first make a pre-
request in order to find out whether they will get a positive response to their
request. Levinson describes this in an underlying structure consisting of four
positions. Below is an example and the underlying structure.
(20) A: (1) Are you doing anything special tonight?
B: (2) No, not really. Why?
A: (3) Well, I wanted to ask if you would like to go out to
dinner with me.
B: (4) I’d love to.
Pragmatics
The underlying structure of (20):
(1) Pre-request
(2) ‘Go ahead’ reaction
(3) Request
(4) Consent
Goffman’s work on ‘face’ offers an explanation for the pre-
request phenomenon. If B had given an evasive answer to the
pre-request, then that would have eliminated the necessity of
making the main request, preventing the loss of face of either
participant. A does not have to deal with a refusal and B does
not have to refuse the request in a direct manner; after the pre-
request, B can claim to be extremely busy which will soften
the blow of the refusal.
Pragmatics
Indirect requests have certain similarities with pre-requests in that both are attempts to
ascertain whether or not there are grounds for refusing a direct request. Consider the
following example. A customer walks into a shoestore and asks:
(21) Do you sell jogging shoes?
This question is actually a preliminary check to see if the sales clerk will be able to
give an affirmative response to a request to see an assortment of jogging shoes. In
Levinson’s opinion, indirect requests can be viewed as pre-requests in an underlying
structure consisting of four positions.
(22) A: (1) Do you sell jogging shoes?
B: (2) Yes.
A: (3) Would you show me some, please?
B: (4) I’ll go get them for you.
In many cases the reaction to a pre-request is the same as to the direct request.
(23) A: (1) Do you sell jogging shoes?
B: (2) Yes, I’ll show you some.
A: (3) Thank you.
This reduction can be explained with the politeness strategy. It ensures that the
customer does not lose face; the customer is no longer obliged to formulate a direct
request.
Pragmatics
Presupposition
• A presupposition is something the speaker assumes
to be the case prior to making an utterance.
• Presuppositions are inferences about what is assumed to be true in
the utterance rather than directly asserted to be true (Peccei, J.
1999).

e.g. Mary’s brother bought three horses.

In producing the above utterance, the speaker will
normally be expected to have the presuppositions that
• a person called Mary exists;
• she has a brother;
 Constancy under negation
Basically, it means that the presupposition
of a statement will remain constant (i.e. still true)
even when that statement is negated.
e.g. Mary’s dog is cute presupposes that
Mary has a dog. Mary’s dog isn’t cute
also presupposes that Mary has a dog.
• Presupposition vs. entailment
e.g. John managed to stop in time.
From the example we can infer:
John stopped in time.
John tried to stop in time.
Now take the negation of the example:
John didn’t manage to stop in time.
From this we cannot infer “John stopped in time”.
yet the inference to “John tried to stop in time” .
is preserved and thus shared by the example
sentence and its negative form.
• On the basis of the negation test, we identify
that the presupposition of both the example
sentence and its negative form is “John tried to stop in
time”, but not “John stopped in time”.

Clearly, the example sentence entails

“John stopped in time”, but its negative form
does not entail it. Negation alters a sentence’s entailment,
but it leaves the presuppositions
untouched.
• A statement A presupposes another
statement B iff:
1) if A is true, then B is true.
2) if A is false, then B is true.
e.g. John is married.
John exists.
John is not married.
John exists.
• A statement A entails another statement B iff:
1) if A is true, then B is true.
2) if A is false, then B may or may not be true.

e.g. 1) That person is a bachelor.

That person is a man.

2) That person is not a bachelor.

That person can be a man or a woman.
Types of presupposition
• Existential presupposition
• Factive presupposition
• Lexical presupposition
• Structural presuppositions
• Non-factive presupposition
• Counter-factual presupposition
Existential presupposition
The existential presupposition is present in
possessive constructions as well as any definite
noun phrase.

X: Your car
Y: You have a car.
• Your car >> You have a car.

the King of Sweden, the cat , the girl next door

Factive presupposition
• The presupposed information following a verb or verbal phrase
like know, realize, regret, be aware, be odd, be glad, etc. can be
treated as a fact, and is described as a factive presupposition.
• A number of verbs or verbal phrases such as know, realize,
regret, be aware, be odd, be glad, etc. have factive
presuppositions.
e.g. She didn’t realize he was ill >>He was ill.
We regret telling him >> We told him.
I wasn’t aware that she was married >> She was married.
It isn’t odd that he left early >> He left early.
I’m glad that it’s over >>It’s over.
Lexical presupposition
• In Lexical presupposition, the use of one form with
its asserted meaning is conventionally interpreted
with the presupposition that another (non-asserted)
meaning is understood.
e.g.Each time you say that someone managed to do something
the asserted meaning is that the person succeeded in some way.
When you say that someone “didn’t manage”, the asserted
meaning is that the person didn’t succeed. In both cases,
however, there is a presupposition (non-asserted) that the
person “tried” to do that something. So, “managed” is
conventionally interpreted as asserting “succeeded” and
presupposing “tried”
• Other lexical items include: stop, start, and again
e.g.
• He managed to do it >> He tried to do it.
• He stopped smoking >> He used to smoke.
• They started complaining >> They weren’t
complaining before.
• You are late again >> You were late before.
• In the case of lexical presupposition, the speaker’s
use of a particular expression is taken to presuppose
another (unstated) concept, whereas in the case of a
factive presupposition, the use of a particular
expression is taken to presuppose the truth of the
information that is stated after it.
Structural presuppositions
• Certain sentence structures have been analyzed as
conventionally and regularly presupposing that part
of the structure is already assumed to be true.
e.g. When did he leave ? (>> He left.)
Where did you buy the ticket ? (>> You bought
the ticket.)
• The wh-question construction in English, as shown
above, is conventionally interpreted with the
presupposition that the information after the wh-form
is already known to be the case.
Non-factive presupposition
• So far, we have only considered contexts in which
presuppositions are assumed to be true. There are,
however, examples of non-factive presuppositions
associated with a number of verbs in English.
• A non-factive presupposition is one that is assumed
not to be true.
• Verbs like dream, imagine, and pretend are used
with the presupposition that what follows is not true.
• I dreamed that I was rich. >> I was not rich.
• We imagined we were in Hawaii. >> We were not in
Hawaii.
• He pretends to be ill. >> He is not ill.
Counter-factual presupposition
What is presupposed is not only not true, but is the
opposite of what is true, or “contrary to facts”.
• A counterfactual conditional structure presupposes
that the information in the if-clause is not true at
the time of utterance.
• If you were my friend, you would have helped me.
>> You are not my friend.
Type Example Presupposition

Existential the X >> X exists

Factive I regret leaving >> I left
Non-factive He pretended to be happy >> He wasn’t happy
Lexical He managed to escape >> He tried to escape
Structural When did she die? >> She died
Counterfactual If I weren’t ill, >> I am ill
A property of presupposition
• Defeasibility
In some contexts, a presupposition is defeasible.
1. Linguistic factors
e.g. John didn’t manage to pass the exam. >> John tried
to pass the exam.
But if uttered in the context,
John didn’t manage to pass the exam. In fact he didn’t
even try to.
then the original presupposition is cancelled by the
utterance In fact he didn’t even try to.
e.g. It won’t be Luke who will betray you.
>> Someone will betray you.
But if uttered in the context,
You say that someone in this room will betray
you. Well, maybe so. But it won’t be Luke who
will betray you. It won’t be Paul. It won’t be
Bill, and it certainly won’t be John. Therefore no
one in this room is actually going to betray you.
then the original presupposition is cancelled by
the last sentence no one in this room is actually
going to betray you.
2.Non-linguistic factors
e.g. She cried before she finished writing the novel. >>
She finished writing the novel.
But in the utterance,
She died before she finished writing the novel.
the presupposition She finished writing the novel is
cancelled by the common sense knowledge that if
someone dies, he cannot write novels anymore.
• The projection problem

• It was originally suggested that presuppositions of

complex sentences are equal to the presuppositions of
sentence 1 plus the presuppositions of sentence 2 plus
the presuppositions of sentence n. But such a simple
solution is far from correct. This compositional
problem is known as the projection problem for
presuppositions.
• E.g.

a. John didn’t manage to pass his exams.>> John tried to pass his
exams.
b. John didn’t manage to pass his exams, in fact he didn’t even try
to.
c. John didn’t manage to pass his exams, if indeed he even tried.
d. Either John never tried to pass his exams, or he tried but he
never managed to pass them.
e. John didn’t manage to pass his exams; he got through without
even trying.
In b, c, d, e, the presupposition of the first clause “John tried to pass
his exams” has dropped out just because of the propositions
expressed by the rest of the clauses of the complex whole.
Small differences, big difference:
Interactional Sociolinguistics

• ‘Small stuff’ can make a big difference---communication may

succeed or fail because of it. This approach highlights the
importance of small and subtle variations in the way people
use and interpret spoken discourse.
Interactional Sociolinguistics

• What Is Interactional Sociolinguistics?

Interactional Sociolinguistics has connections
with other approaches to spoken interaction that
we have already examined (especially the
ethnography of speaking), but as its name
suggests, it also has connections with
sociolinguistics, the study of socially
conditioned patterns of variation in language-
use.
Sociolinguists describe differences in the way
people use language---their pronunciation, their
grammar, the words they use for particular
things, the way that bilingual speakers alternate
between two or more languages---and seek to
explain these differences by correlating them
with nonlinguistic differences---for instance
with speakers’ class, race, gender, the formality
of the situation, the subject people are talking
about or the setting where they are talking.
Interactional sociolinguists take a similar approach to
phenomena which are important in organizing spoken
interaction. It is not only in their grammar, vocabulary and
pronunciation that language-users have different ways of
saying the same thing (e.g. referring to a carbonated soft
drink as pop or soda or juice, pronouncing or not
pronouncing an /r/ in the word farm). Such aspects of
interaction as turn-taking rules, conventions for indicating
acknowledgement and agreement, the marking of
utterances as particular kinds of speech acts or as
containing important information are also ‘variables.
Interactional sociolinguistics is the study of this kind of
variation.
 As with many of the regularities studied by CA
practitioners, the aspects of interaction that interest interactional
sociolinguists are often ones that the participants in talk have
little or no
conscious awareness of. They could not explicitly explain how
they recognize a certain intonation pattern as signaling
important new information, or why one minimal response (a
brief acknowledgement of another speaker such as mhm, yeah,
right) sounds encouraging, while another sounds dismissive.
They may also be unaware that these aspects of interactional
behaviors are variable, with different meanings for different
groups of language users.
 The nature and significance of this kind of variation may be
illustrated by an example which concerns nonverbal
communicative behavior. Some British schoolchildren of
African-Caribbean ethnicity have been observed to direct their
gaze downwards when confronted by a teacher. This behavior
infuriates some white teachers, prompting them to issue
commands like ‘look at me when I’m talking to you!’. These
teachers understand the children’s gaze behavior as evasive and
disrespectful. For them, ‘looking someone in the eye’ is a mark
of attentiveness and of honesty. However, in the children’s own
community a different assumption is operative: lowering one’s
gaze is a way of conveying respect.
Interactional Sociolinguistics

This is an example of ‘intercultural communication’---

communication between people who are operating within
two different cultural systems.------More exactly it is an
example of intercultural miscommunication. While
misunderstanding is always a possibility in any kind of
spoken interaction, it is a particular problem in intercultural
encounters, because of systematic differences in the cultural
assumptions and patterns of linguistic behavior which are
considered normal by those involved. Small differences in
people’s ways of doing things in conversation can lead to
serious misunderstandings, and since the parties often do not
understand what has gone wrong, the problem may not be
identified as linguistic or communicative at all. Instead it
may be taken as evidence that members of group X are just
mean, or stupid, or crazy.
 In the example given above, the child who behaves in a way
s/he considers respectful, and is then accused of disrespect
by a teacher because of it , may well end up feeling both
confused and resentful. Since the communicational
difference that caused the problem is correlated with a
highly salient social difference --- in this case one of
race/ethnicity –the experience, particularly if it is repeated
over time, may contribute to reinforcing each party’s
stereotypical beliefs about the other. If the teacher thinks
African-Caribbean students are badly behaved and
disrespectful, s/he is likely to have that view confirmed by
the child’s behavior. If the child thinks white teachers treat
black pupils unfairly because of racial prejudice, s/he will
find supporting evidence for that proposition in the
teacher’s behavior.
Interactional Sociolinguistics

In communicative encounters between members of different

groups, the smaller the differences are, the greater their
potential to cause problems. If someone addresses me in a
foreign language that I don’t speak , I know immediately
that I don’t understand them, and it is obvious why. If
someone speaks to me in an unfamiliar dialect of a language
I do speak, and I have difficulty decoding their
pronunciation, the source of my problem will be equally
apparent to me: accent is something we pay attention to, and
have a certain amount of explicit knowledge about.
 Something like the gaze behavior variation described above
is more subtle: teachers have no difficulty in pinpointing
which aspect of the children’s behavior is at odds with their
expectations, but they do not realize that the behavior in
question is not intended to mean what they think it means.
Unlike the standard English speaker who is addressed in
Japanese or Chinese, the teacher may not be aware that s/he
has not understood the student. The proposition that people
speak different languages or pronounce words differently
because they are from different places is common
knowledge; the proposition that their gaze behavior varies
for similar reasons is not.
 There are other cases of interactional variation in which the
participants become aware that they are not understanding
or being understood, but they cannot identify the variables
which are causing the difficulty, because those variables are
features of discourse which they do not attend to
consciously. For instance, prosodic and paralinguistic
features are not represented in written language in any
precision, and unlike spelling, punctuation or grammar, they
are not explicitly taught. As a result, people tend to lack
awareness about the important and systematic contribution
they make to the meaning of spoken discourse .
•
Interactional Sociolinguistics

• Contextualization cues and ‘crosstalk’

Some important research in interactional
sociolinguistics has been conducted with a view to
helping people who regularly engage in
intercultural communication (such as teachers or
workers in multi-ethnic schools or businesses )
become aware of the differences that may cause
problems, and take account of variation in their
real life encounters with speakers whose ways of
interacting differ from their own.
An example is the work of John Gumperz and his
associates (Gumperz et al. 1979) on ‘crosstalk’
between members of Britain’s white majority
community and members of Asian minority
communities. Communication problems involving
Asian speakers are sometimes attributed by white
Britons to the fact that Asians are ‘foreigners’ and
have difficulty with the English language. But while
some Asians may indeed have a limited command of
English, others, including many who were born in
India or Pakistan, speak it fluently as a second
language.
However, what such people speak is not necessarily
British English but an English variety developed in
the Indian subcontinent itself. In Britain this
becomes an ‘ethnic’ variety, a mark of membership
in a particular minority ethnic community.
Gumperz and his associates set out to investigate
what systematic differences between this variety
and the majority variety of English might be
causing the perception of ‘communication
problems’ between white and Asian speakers.
 Interactional Sociolinguistics
• An example discussed by Gumperz (1982) comes from a
workplace in which white workers ate their lunches in a
cafeteria staffed mainly by Asian workers. A customer
who had chosen meat was addressed by a cafeteria
worker who uttered the word gravy, pronouncing the
word with falling intonation. The function of this
utterance was to ask whether the customer wanted gravy ,
but it was not heard as an offer or a question. Rather it
sounded like an assertion: ‘this is gravy’ or ‘I’ m giving
you gravy –which seemed rude and unnecessary, since
customers could see for themselves what it was and
decide for themselves whether they wanted any.
• Gumperz points out that British English speakers
subconsciously expect offers of this kind to be
pronounced with rising intonation (gravy?).When
Asians did not produce the expected pattern, white
workers inferred that they were deliberately being rude.
But in Indian varieties of English, falling intonation on
offers has the same meaning as rising intonation in
British varieties---in other words, there is a systematic
difference in the conventions used by the two groups
for indicating the status of utterance as an offer. Since
neither group was explicitly aware of that difference,
the result in his case was misunderstanding.
Interactional Sociolinguistics
• The use of rising or falling intonation to signal ‘this is an
offer’ is an example of what Gumperz calls a
‘contextualization cue’. When we speak we have ways of
conveying to the listener some quite complex information
about how we intend them to treat the message: “I am
speaking to you confidentially as a trusted friend ”, or “I
intend this ironically, not seriously”, or “this is my main
point , so listen carefully ”. The means by which we
communicate this kind of contextualizing information
include prosody (intonation, pitch and stress contrasts ),
paralinguistic cues (hesitation, pausing ,contrasts of speed),
or switching to a different language, dialect , style or register
for the part of the utterance we want to pick out as having a
particular significance.
• If I am talking to someone and I do not understand
their contextualization cues, I will miss part of the
meaning they are trying to communicate. For instance,
if I don’t recognize that someone’s tone of voice
indicates irony , I may end up thinking they mean
something different or even opposite from what they
intended. If I do not pick up on the fact that speaking
very softly and rapidly is meant to signal ‘this is
confidential or potentially damaging information I may
not infer that I should be careful about repeating it to
others. If I cannot distinguish important from less
important information using contextualization cues I
will have trouble following the thread of the argument
my interlocutor is making, and may wonder, ‘what is
the point here?’.
 As a further illustration of the workings of
contextualization cues, consider the following transcript of
a role-play exercise devised for use in industrial language
training (it is reproduced in Roberts et al. 1992:151). The
scenario involves a worker who wants to retrieve some
timesheets from a colleague. In this transcript speaker A
speaks Indian English and speaker B British English.

• Transcription conventions: // major boundary; / minor

boundary; underlining stress; ↑rise in pitch; ? rising
intonation
Interactional Sociolinguistics

• 1. A: excuse me David//
• 2. B: yeah
• 3. A: could I have those sheets back please//
• 4. B: those sheets?/ I put on your desk yesterday//
• 5. A: haven’t seen on my↑desk//
• 6. B: ah/ I’m sure I did/ I’m sure I did.
• 7. A: no/no/ I’ve not seen on my desk/you’re/are you sure?
• 8. B: absolutely certain/ it said on the top/ checklist for April// those
9. are the ones aren’t they//
10. A: no/no/ not checklists/ I want the timesheets//
11. B: oh the timesheets//sorry/yes/I’ve got them here//
• Trainees who hear a tape of this exchange---and also a
tape in which the same scenario was role-played by
two Anglo-British speakers--- commented that the two
speakers lacked rapport and that the Indian English
speaker sometimes sounded ‘extremely angry’ or
‘slightly offensive’.
Interactional Sociolinguistics

• In this role-play, the negative impression commented on by

trainees is produced to a large extent by the way speaker A
uses stress, which is different from the British English way
of using it. For instance, at line 3, A says ‘could I have
those sheets back please’, where the unmarked British
pattern would be ‘could I have those sheets back please’.
To the British ear, stressing back makes the request sound
like an accusation---‘you took my sheets and you should
give them back’. Then at line 10, when A has realized what
the problem is---B is talking about checklists whereas A’s
request was for timesheets---A stresses lists in the word
checklist, and puts no stress on the item which contains the
really important information, timesheets.
• A British English speaker would probably use contrastive
stress to clarify the point at issue here--- something like ‘no
not the checklists, the timesheets’. The Indian English
speaker’s use of a different stress pattern makes the
message confusing to the British hearer, who may wonder
why lists is being emphasized. The impression of
impoliteness conveyed by stressing back in ‘could I have
those sheets back please’ is reinforced by the Indian
English speaker’s rise in pitch at line 5 and his use of ‘no,
no’ when B says ‘I’m sure I did [put them on your desk]’.
Rising pitch is associated by British English speakers with
excitement or annoyance, while saying ‘no’ in this
context may be heard as a suggestion that B is lying when
he claims to have put the sheets on A’s desk.
• Miscommunication can result not only from variation in
the use of contextualization cues, but also from
conflicting assumptions about the norms and
conventions of particular speech events. Roberts et al.
(1992) cite cases where lack of familiarity with local
assumptions has led Asian candidates at job interviews to
make incorrect inferences about the interviewer’s intentions
in asking certain questions. The following is an example.

• A: What have you been doing since you were made redundant?
• B: Nothing.
• The question is actually a veiled invitation to the
candidate to reassure the selectors that s/he has
not simply been doing nothing since losing a
previous job. But since this is not made explicit
in the wording of the question, understanding it
depends on sharing, or at least being aware of,
certain culturally specific ideas about work and
the lack of it.
Interactional Sociolinguistics
• In the culture the interviewer belongs to, unemployment is often
regarded less as a misfortune than as a moral failing--- a sign
that the unemployed person does not have the necessary ‘work
ethic’( 职 业 道 德 ). Where a job applicant is not currently
employed, s/he will often be expected to provide evidence that
s/he is not merely idle and apathetic. But speaker B in the above
exchange fails to understand this; he takes it that the interviewer
is concerned simply to elicit facts about his present situation,
and answers accordingly. The example shows that ‘crosstalk’
between people of differing cultural backgrounds is not just
a matter of surface linguistic features but also of the
assumptions language-users make about the kind of speech
event they are participating in and what is appropriate or
‘normal’ in that context.
5. Ideology: critical discourse analysis （ CDA)

CDA sees discourse---language use in speech and

writing---as a form of ‘social practice’. Describing
discourse as social practice implies a dialectical
relationship between a particular discursive event and
the situation (s), institution (s) and social structure (s)
which frame it. A dialectical relationship is a two-way
relationship: the discursive event is shaped by
situations, institutions and social structures, but it also
shapes it.
 CDA

• Discourse and social change

During the 1980s and 1990s, a number of critical discourse
analysts, notably the British linguist Norman Fairclough,
turned their attention to the study of ‘language and
power’ ,‘language and ideology’ and ‘language and social
changes’.
The 1980s and 1990s have been analyzed as a time when
the social and political order that had been established after
the Second World War was radically and irrevocably
changed.
The ending of Cold War between the two global superpowers
( the US and USSR) coincided with, and enabled the spread
of, a new political consensus based on liberal democracy and
free market capitalism. Together with the ‘information
revolution’ made possible by new computer technology, the
removal of restrictions on free movement of capital led to the
development of a global market system beyond the control of
any one nation state. In many western countries, the free
market also became a model for non-capitalist institutions
such as hospitals, schools and universities. In Britain, for
example, the National Health Service (NHS) began to be
managed as if it were a business---not to make a profit, but
because of an assumption that all organizations work better
when run on market principles. Users of pubic services were
increasingly treated as if they were ‘customers’.
 CDA
• What has any of this to do with discourse
analysis? One answer is that social changes of
this kind are bound to manifest themselves in all
areas of social life, including the way people
communicate; thus studying the discourse in
which people communicate during a period of
major social change is one way of studying
change itself.
• Another answer is that if an institution wants to change the
relations it has traditionally had with people---for instance,
to recast hospital patients or university students as
‘customers’ ---then one of the things it has to do is instruct
them in their new roles and relationships by changing the
way it addresses them. Not only does this put them in the
position of ‘customers’ in particular exchanges, arguably it
has the more general effect of encouraging them to think
about hospitals or universities in a ‘consumerist’ frame,
while discouraging the use of an older, ‘public service’
frame. From this point of view, the emergence of new
kinds of discourse is not only a consequence of social
change, but also an instrument of social change.
CDA
• Norman Fairclough has argued that a number of
general tendencies representing social change can be
discerned in contemporary institutional discourse. One
of these, which goes along with the adoption of the
capitalist free market as a model for all kinds of
transactions, is a tendency for discourse genres which
were once primarily ‘informational’ to become more
‘promotional’--- they are no longer designed simply to
‘tell’, but also to sell. One salient way in which the
shift is expressed is through the incorporation of
features from the genre of commercial advertising into
non-commercial genres.
• An example Fairclough has examined by comparing
earlier and later texts is the British university
prospectus. Recent prospectuses have much in common
with advertising brochures in their appearance and the
language they use: the reader deciding where to study
is now positioned as a ‘customer’, for whose business
the university must compete by drawing attention to its
‘selling points’. Earlier prospectuses read very
differently. They inform readers about the courses on
offer, the admissions requirements, and so on, but pay
less attention to framing this information in ways
designed to appeal to the reader as a consumer.
CDA

Racist discourse

The Dutch critical discourse analyst Teun van Dijk has

examined the representation of race in mainstream
sources including press reports, speeches made in
European legislative assemblies, school textbooks,
scientific and corporate discourse (van Dijk 1987, 1991,
1996 ).
One issue that is significant is the labeling of the groups under
discussion. It is characteristic of discourse on race that
discussion is framed in a language of ‘them’ and ‘us’, with
minority ethnic groups often marked as outsiders by using
terms like immigrants and foreigners. In some cases these
descriptions are factually accurate, but in many they are not,
suggesting that facts are not the main consideration influencing
the choice of label. The European-born children of immigrants
, for example, are not themselves immigrants, since they did not
migrate from anywhere. Referring to them as such conveys
That unlike others who were born in a certain country, they do
not really ‘belong’ there.
CDA

• The following extract from a report in the British popular

newspaper the Sun illustrates a number of commonplace
features of racist discourse (van Dijk 1996:98)
BRITAIN INVADED BY AN ARMY OF ILLEGALS
Britain is being swamped by a tide of illegal immigrants so desperate for a job
that they will work for a pittance in our restaurants, cafes and nightclubs.
Immigration officers are being overwhelmed with work. Last year, 2191 ‘illegals’
were nabbed and sent back home. But there were tens of thousands more, slaving
behind bars, cleaning hotel rooms and working in kitchens…
Illegals sneak in by:
●DECEIVING immigration officers when they are quizzed at airports.
●DISAPPEARING after their entry visas run out.
●FORGING work permits and other documents.
●RUNNING AWAY from immigrant detention centers.
Source: Sun, 2 February 1989
CDA

• This report uses two metaphors which van Dijk found

to be common in discourse on the subject of
immigration: immigrants are figured as an invading
army or a ‘tide’ threatening to ‘swamp’ the native
population. The advent of migrants is thus likened to a
military incursion or a natural disaster, which also
suggests that fear and resistance are reasonable
responses. Two other common features reproduced by
the Sun are high but vague numbers (undetected illegal
immigrants are said to number ‘tens of thousands’ ) and
the attribution of dishonest or criminal behavior to the
group under discussion (cf. the verbs sneak, deceive,
disappear ,forge ,run away ).
•
• As van Dijk notes, there are some surface indications of
sympathy for the plight of the illegal immigrants, who are
described as ‘desperate’ , ‘working for a pittance (a very small
amount of money)’, and ‘slaving ’. However, he goes on to
make the important point that what is actually communicated
by a text may not be only or exactly what is explicitly said in
that text. As we know, communication involves inference, and
meaning is produced by placing new information in the
context of existing background knowledge and established
frameworks for interpretation.
• Here the relevant framework is one in which
immigrants/foreigners pose a threat to the economic
wellbeing of native workers because they are
sufficiently ‘desperate’ to work ‘for a pittance.’ In
other words, ‘they take our jobs’. This idea is
sufficiently familiar from a whole history of racist
discourse that the report does not have to spell its
relevance out: it can be ‘cued ’by a couple of
significant lexical choices like pittance and slaving.
CDA

Van Dijk points out that press reports on the subject of

illegal immigration very rarely make any reference to
the employers who knowingly take on immigrants to
work illegally, using their economic desperation and
precarious legal position to exploit them by paying ‘a
pittance’. Logically, the actions of these employers are
as much a part of the ‘problem’---and just as much
against the law –as the actions of the immigrants
themselves. But in press reports the immigrants are
consistently foregrounded while the employers are
backgrounded.
• The Sun article quoted above, for instance, goes on to
report that in one raid, immigration officials ‘ended up
taking away 13 Nigerians, all employed illegally’. The
grammar of this sentence deletes the agent of the verb
employ, focusing on the Nigerians rather than the
(presumably British )business people who illegally
employed them.
6. Variation Analysis

This approach stems largely from studies of variation and

change in language: fundamental assumptions of such
studies are that linguistic variation is patterned both
socially and linguistically, and that such patterns can be
discovered only through systematic investigation of a
speech community. Thus, variationists try to discover
patterns in the distribution of alternative ways of saying
the same thing, i.e. the social and linguistic factors that
are responsible for variation in ways of speaking.
 Both the initial methodology and the theory underlying
such studies are those of William Labov. It is in the
search for text structure, the analysis of text-level
variants and of how text constraints other forms, that a
variationist approach to discourse has been developed.
The origins of the variationist approach are solely
within linguistics. Variation theory focuses on structural
categories within texts and the way syntactic structure
(and variation) helps to define and realize those structures.
 Labov(1972), Labov and Waletsky(1967)provided a
systematic framework for the analysis of oral narrative-
a framework that illustrates quite well the variationist
approach to discourse units. This framework defines a
narrative as a particular bounded unit in discourse, and it
defines parts of narrative as smaller units whose identities
are based on their linguistic (syntactic, semantic) properties
and on their role in the narrative.
 Narratives have a linear structure in which different
sections present different kinds of information. Each
section has a different function within the story. In
addition, each section is comprised of different types of
clauses whose syntactic and semantic properties
contribute to their identity as units within the story, and
to their function.
Narrative structure
Narratives are opened by an abstract, a clause that
summarizes the experience and presents a general
proposition that the narrative will expand. Orientation
clauses follow the abstract: they describe background
information such as time, place, and identity of characters.
The main part of the narrative is comprised of
complicating action clauses. Each complicating action
clause describes an event--- a bounded occurrence in
time--- that is understood to shift reference time, i.e. it
follows the event immediately preceding it, and precedes
the event immediately following it.
Evaluation pervades the narrative: speakers can comment
on events from outside of the story world, suspend the
action through embedded orientation clauses, and report
events that themselves indicate the significance of the
experience. Result or resolution is concerned with what
finally happened. Finally, the story is closed by a coda---
a clause that shifts out of the past time frame of the story
to the time frame of the current talk.
1.Abstract: What, in a nutshell, is this story about?
2.Orientation: Who, when, where, what?
3.Complicating action: Then what happened?
4.Evaluation: So what, how is this interesting?
5.Result or resolution: what finally happened?
6.Coda: That’s it, I’ve finished and am “bridging”
back to our present situation.
Variable relationships in discourse

• Although the narrative is an example of a textual unit, it is

important to note that variationists typically focus upon other units
of analysis that are not text sized. As Labov (1972)observes: It is
common for a language to have many alternate ways of saying
“the same” thing. e.g. working vs. workin’
One of the main tasks in variation analysis is to discover
constraints that ( can be linguistic and /or social) help determine
which realization of a single underlying representation appears in
the surface form of the utterance. Consider the alternation between
working vs. workin . It is constrained both linguistically and
socially. The use of in’ is more frequent in progressives and
participles than in gerunds---reflecting different patterns of
historical development. In’ is also more frequently used by working
class than middle class speakers and by men than women;
all speakers use in’ more than ing when they are speaking casually
instead of carefully.
Since variationists try to discover patterns in the distribution of
alternative ways of saying the same thing, i.e. the social and
linguistic constraints on linguistic variation, an initial step in
variationist studies is to establish which forms alternate with one
another and in which environments they can do so. Constraints
are then proposed in line with different hypotheses that may
propose a variety of different explanations for a given
phenomenon, ranging from the physiology of articulation
to processing considerations to social or biological
universals. Variationist approaches compare different
explanations searching for data that confirm (or cast
doubt on) the co-occurrences predicted by each
explanation. We have seen in this section that an
important unit of analysis for variationist is the linguistic
variable.
Other approaches
7. Systemic Functional Grammar
M.A.K. Halliday
8. Relevance Theory
D. Sperber & D. Wilson
Collecting and Transcribing Data
Linguistic data
Types of linguistic data
• Linguistic data: intuitive data 直觉语料
non-intuitive data
• Intuitive data: introspective data （ the
data created by researchers ） （ on chair ）
elicitation data
• Non-intuitive data: anecdotal data
corpus data 语料库
Collecting Data
• If you are going to analyze any kind of discourse, you
need material (‘data’) to work on. Collecting data involves
recording speech, gathering written texts, obtaining
permission from speakers and writers to use their texts, and
keeping careful records about the texts collected and the
individuals from whom they were obtained.
• Compared with collecting spoken language data, most
written materials are relatively straightforward to collect.
• Spoken language data for discourse analysis consist in the
first instance of recordings (audio or video) of people
talking; the next part of the process is to construct a
transcript, a representation of talk in written form which
will serve as the main input to analysis.
Collecting Data
• Speech is the primary mode of human
communication. As a result, all kinds of speech
have evolved: there are not only spontaneous
multi-party dialogues but scripted and unscripted
monologues, radio and television interviews,
telephone conversations, class lectures, and so
forth.
• There are some methodological considerations
that must be addressed in the collection of speech.
In collecting any kind of speech, the central
concern is obtaining speech that is ‘natural’. This is
particularly important when gathering spontaneous
multi-party dialogues, such as informal talks
between two or more individuals. If multi-party
dialogues are not carefully collected, the result can
be a series of recordings containing unnatural
speech.
• Collecting ‘natural’ multi-party dialogues involves more
than simply turning on a tape recorder and having people
converse. As anyone who has collected language data
knows, as soon as you turn on a tape recorder and tell
people to speak, you have to deal with the common
scientific problem known as the ‘observer’s paradox’: by
observing something, you change its natural behavior. And
this is especially true with speech because many individuals
are quite insecure about the way that they speak, and when
they know that their speech is being monitored, they may
adjust their speaking style and attempt to speak what they
perceive to be ‘good’ Chinese or English or any other
language.
Collecting Data
• To avoid this problem, those creating earlier corpora
recorded people surreptitiously, and only after recordings
were secretly made were individuals informed that they had
been recorded. While it may have been acceptable and legal
back in the 1950s and 1960s to record individuals without
their knowledge, now such recordings are illegal in some
places and considered unethical by most linguists. It is
therefore imperative that anyone recording speech not only
informs individuals that they are being recorded but obtains
written permission from the individuals to use their speech
in the corpus.
•
• Since it is not possible to include surreptitious
speech in a corpus, does this mean that non-
surreptitiously gathered speech is not natural?
This is an open question, since it is not possible to
answer it with any definite certainty: so little
speech has been collected and included in corpora
that it is not really possible to know empirically
precisely what natural speech is really like.
However, there are ways to increase the
probability that the speech included in a corpus
will be natural.
Collecting Data
• First, before individuals are recorded, they should
be given in written form a brief description of the
project they are participating in. In this description,
the purpose of the project should be described, and it
should be stressed that speech is being collected for
descriptive linguistic research, not to determine
whether those being recorded are speaking ‘good’ or
‘bad’ Chinese or English or whatever.
• A second way to enhance naturalness is to record
as lengthy a conversation as possible so that when
the conversation is transcribed, the transcriber can
select the most natural segment of speech from a
much lengthier speech sample. The initial part of a
conversation can be discarded, since people are
sometimes nervous and hesitant upon first being
recorded but after a while become less self-
conscious and start speaking more naturally.
Collecting Data
• When it comes time to actually make recordings,
whoever is making the recordings needs to follow a few
basic principles to ensure the most natural recordings
possible.
• As far as possible, those making recordings should try to
record individuals in the actual environments in which
they typically speak. Because the goal is to record people
in natural speaking environments, it is best for the research
assistant not to be present during recordings. He or she can
simply set up the recording equipment, turn on the
recorder, and leave. Alternatively, the people being
recorded can be loaned the recording equipment and taught
to set it up and turn on the recorder themselves.
•
• While collecting natural speech is a key issue
when recording multi-party dialogues, it is less of
an issue with other types of speech. For instance,
those participating in radio and television
broadcasts will undoubtedly be conscious of the
way they are speaking, and therefore may heavily
monitor what they say. However, heavily
monitored speech is ‘natural’ speech in this
context, so this is precisely the kind of speech one
wants to gather.
• In recording any kind of speech, the type of
recorder to use is an important issue to consider.
With the rise of digital technology, there are now
digital tape recorders that can make high-quality
digital recordings.
• When working with spoken recordings, it is often
desirable to work with digital recordings on a
computer to prepare the recording for inclusion on a
CD for storage or distribution.
• Advantages and disadvantages of tape-recorded data (rather
than observational data)
• As Erikson (1986)points out, you cannot interact with a
tape unless you are part of the interaction being taped, you
cannot ,for example, stop people and ask them why they
said what they just did, or get them to repeat something that
is not clear. In addition, contextual information that might
be important may be unavailable. Audiotape does not reveal
whether people were looking at one another or how their
bodies were aligned; even videotape shows only what the
camera was aimed at. It is, of course, possible to alleviate
these problems by being there when the tape is being made
and taking good notes, but even so there are bound to be
questions that arise later.
• On the other hand, Erikson says, tape-recording can help
make an analysis more complete and more fine-grained
than purely observational approach would. A tape you can
hear or see over and over makes it possible to notice and
take seriously things that happen only once or rarely. In
real-time observation we tend to notice things that are
repeated. And a tape can help keep the analyst from
jumping to conclusions that may be reinforced by selective
memory. Baugh(1993)points out another advantage of tape-
recording: the researcher does not have to be the one to
collect the data, if he or she is not the ideal fieldworker for
the situation.
Collecting Data
• Other data sources
• There are some kinds of spoken data which
you don’t have to record (or ask informants to
record) because they already exist and in the
public domain. An obvious example is talk that
is broadcast on radio or television.
• The new medium of the internet is another potential
source of data, in two ways. First, communication on the
net--- on discussion lists, bulletin boards, chat rooms---is
an object of considerable interest to discourse analysts,
who point out that although the medium is writing, it has
many ‘speech-like’ features. Second, there are sites where
a user can download linguistic data that has been
collected by other researchers. A collection of linguistic
data is known as a corpus, and the availability of more
and more corpora on-line is an important recent
development.
Collecting Data
• One obvious drawback with getting spoken
language data from an on-line corpus is that
typically you only get the written transcript.
(Sometimes you can get sound, but it is more likely
to be isolated syllables or words than whole chunks
of conversation.) However, the transcript may
include additional information by way of ‘tagging’,
where the original analyst has coded the data, using
symbols to indicate things like pitch, stress and
volume. If you know the relevant symbol you can
search for the phenomenon it codes.
•
• Whether this kind of data is useful depends on
what you want to investigate. Potentially, there is
an enormous advantage in having access to
corpora far larger than you could collect and
transcribe yourself: it is especially advantageous
if you want a lot of examples of a very well-
defined feature, or if you want to check how
common something is statistically.
Collecting Data
• How much data?
• How much data is enough? It depends on your goals,
your resources, and the kinds of claims you are hoping
to be able to make.
• Some researchers have worked on very small and
specific data samples: Deborah Tannen is famous for
an extremely detailed study of the talk which took
place at a single Thanksgiving dinner (Tannen ,1984).
Jennifer Coates’s study Women Talk: Conversation
between Women Friends(1996) is based on a larger
sample of data from several different groups of women
friends, reflecting the fact that Coates was interested in
similarities and differences among women of different
ages and backgrounds.
• A factor that affects how much data you may have to record
is how frequently the variable you are interested in occurs in
talk. For instance, there is a well known article called “ On
questions following questions in service encounters” (Merritt,
1976). Questions following questions in service encounters
(for example ‘Can I have a pint of Guinness?’ ‘Are you 18?’)
are not exactly rare, but nor do they occur in every service
encounter: a researcher will need to record quite a large
number of encounters to get a reasonable number in which
questions follow questions. By contrast, if you are
researching something like hesitations, or minimal responses
like ‘mhm’, ‘yeah’, you can be confident that they will occur
very frequently in any exchange you record. It will not take
long to collect a respectable number of examples.
 Collecting Data

• While in principle it might seem desirable to

have as many examples as possible, in practice
you have to draw the line somewhere, bearing
in mind that the collection of data is only the
beginning of a lengthy process which includes
transcription as well as the actual analysis.
Transcribing the Data
• Transcription involves capturing who said what,
in what manner (e.g. prosody, pause, voice quality),
to whom, under what circumstances (e.g. setting,
activity, participant characteristics and relationships
to one another). It includes preservation of various
temporal aspects (e.g. pause duration, sequence of
events, and simultaneity or overlap of speaker turns
or speech and gestures), and some metacomments.
• Transcribing speech is in essence a highly artificial
process, since an exclusively oral form of language
is represented in written form. Consequently, before
any transcription is undertaken, it is important to
decide just how much that exists in a spoken text
one wishes to include in a transcription of it. In the
research done in the past, researchers have varied
considerably in how much detail they have included
in their transcriptions of speech.
Transcribing the Data
• Whether one does a minimal or detailed
transcription of speech, it is important to realize
that it is not possible to record all of the subtleties
of speech in a written transcription. As
Cook(1995) notes: a spoken text is made
meaningful by more than the words one finds in a
transcription: how a conversation is interpreted
depends crucially upon such contextual features as
paralanguage (e.g. gestures and facial
expressions), the knowledge the participants have
about the cultural context in which the
conversation takes place, their attitudes towards
one another, and so forth.
• All of this extra-linguistic information is very difficult to
encode in a written transcription without the researcher
developing an elaborate system of annotation to mark this
information and the transcriber spending hours both
interpreting what is going on in a conversation and inserting
the relevant mark-up. It is therefore advisable when
transcribing speech to find a middle ground: To provide an
accurate transcription of what people actually said in a
conversation, and then, if resources permit, to add extra
information (e.g. annotation marking features of intonation)
A few simple steps that will make the transcription easier to
read and refer to:
 1. Use actual, spelled out names for the
speakers, not initials.( These should generally
not be their real names, but rather pseudonyms)
2. Number the lines of the transcript so you can
refer the reader to the precise parts you discuss
in your paper.
3. In excerpts from transcripts that you use as
examples, highlight the relevant lines, words, or
phrases so readers can pinpoint exactly what to
focus on.
Design Your Own Project
Design Your Own Project
• Choosing your research topic
Most ‘good ideas’ for research do not just spring from the
researcher’s imagination, they are suggested by previous
research. Knowledge on any subject is cumulative: people
look at what is already known and notice gaps, or think of
objections or supplementary questions. There are several
ways in which previous research can be used to help design a
new research project.
Design Your Own Project
• Replicate
The simplest kind of research project based on existing research involves
replicating a previous study. You ask the same question(s) and use the
same methods as a researcher whose work you have read, but you collect
your own data: your question is whether your findings will resemble the
original findings. This might sound dull, but in fact it is quite important to
know whether patterns reported in research turn up consistently. Only
when similar results have been obtained across a range of different studies
can we make reliable generalizations about whatever is at issue. Another
reason for replicating old studies is to find out if things have changed over
time.
A variant on straightforward replication is to undertake a study asking the
same questions and using the same methods as a previous study, but
deliberately altering one or more of the nonlinguistic variables( e.g. the
social characteristics of the participants or the setting for their interaction).
Design Your Own Project
• Compare and contrast
Another common type of research project compares two
kinds of talk. E.g. a comparison between spontaneous casual
conversation and soap opera dialogue; a comparison between
face-to-face talk and telephone conversation etc.
• Take issue with a previous claim
Sometimes you may feel intuitively that a claim is
misguided and then try to demonstrate empirically why it is
misguided.
Design Your Own Project
• Describe something new
• You can also construct a worthwhile research project by focusing on a kind
of data that has not been analyzed before, and simply describing its discourse
characteristics. For instance, Kay Richardson(1997) has analyzed the talk that
is heard on ‘shopping channels’ that work like mail order catalogues, offering
goods that viewers can buy over the phone. Shopping channel presenters face
a linguistic challenge: since viewers need to be able to inspect the object on
offer for long enough to decide if they want to buy it, presenters have to talk
about a single object , like a ring or a watch, for much longer than would be
typical in other genres of talk. They have to describe the object, with
particular attention to features the camera might distort. They also have to
make it sound attractive, and constantly repeat information like the price and
the phone number to call if you wish to order. The resulting discourse is very
distinctive.
• How is previous research relevant to a project that involves describing
something new? In two ways. First, it is desirable to investigate whether
anyone else actually has published a description of your chosen genre.
Second, there may be a case for approaching a novel phenomenon on the
model of something that has already been studied.
Design Your Own Project
What makes a workable research topic?
 A well-focused idea can be phrased as a question or a set of
closely related questions. (Your questions may become narrower
in the course of your research, but you must have clear questions
with which to start.)
● You know how you will go about answering the questions, that
is, what your research methodology or methodologies will be.
● You are familiar with the research site.
● You have the time it will realistically take to carry out the
research, analyze the results, and write your report.
● The project requires ways of collecting and analyzing data with
which you are comfortable and at which you are competent.
Design Your Own Project
• Putting it in writing
• Most research papers on spoken discourse will have some variant of the following
structure:
1. An introduction section setting out the researcher’s question and why it is of
interest.
2. A literature review surveying relevant previous work on the subject.
3. A methods section explaining in some detail what the researcher’s data consist of,
how they were collected, what approach was taken to analyzing them and why that
approach was chosen as most suitable for the purpose.
4. A section setting out the results of the project
5. A discussion section in which the researcher presents possible explanations for the
results and discusses their meaning or significance.
6. A conclusion in which the researcher summarizes what has been learnt from the
project, what its problems and limitations are, and what questions arise from it for
future research.
7. A References section listing all the resources the researcher has consulted
8. An appendix setting out transcription conventions used by the researcher.
9. An appendix containing transcribed data.
Case Studies (omitted)

• Media discourse; classroom discourse; academic discourse;

children’s discourse; gender discourse; forensic discourse;
workplace discourse; discourse grammar; discourse
comprehension; discourse structure; discourse makers; genre;
reference; cohesion; coherence; etc.
Term paper
Thank you !