Análisis del Discurso en Lengua Inglesa

GRADO EN ESTUDIOS INGLESES: LENGUA, LITERATURA Y CULTURA

Unit 1: a) Introducing Discourse Analysis

b) The data for Discourse Analysis

Professor: Dr. Laura Alba Juez

[email protected]
a) INTRODUCING DISCOURSE ANALYSIS

Defining text and discourse:

What is Text Linguistics? What is Discourse Analysis?
• The terms text and discourse are used in a
variety of ways by different researchers.

• In everyday use, the notion of text has been

limited to written language, whereas
discourse can be said to be used more in
relation with the spoken language.

• However, modern linguistics has introduced

a concept of text that includes every type of
utterance, from magazine articles or
cooking recipes to television interviews or
normal conversations among friends, just to
name a few.
De Beaugrande and Dressler’s (1981) definition of text
De Beaugrande and Dressler take a wider perspective on the concept
of text and define it as a communicative event that must satisfy the
following seven criteria:

1. COHESION, which refers to the close relationship between

text and syntax. E.g.: ellipsis, anaphora, cataphora,
recurrence or conjunction are crucial phenomena for
cohesion.

2. COHERENCE, which has to do with the meaning of the text

and with elements of knowledge or cognitive structures
which are implied by the language used and thus influence
the reception of the message.

3. INTENTIONALITY, which has to do with the attitude of the

speaker/writer.
4. ACCEPTABILITY, which concerns the preparation of the
reader/hearer to assess the relevance or usefulness of a text.

5. INFORMATIVITY, which refers to the quantity and quality of

the information.

6. SITUATIONALITY, which indicates that the situation in which a

text is produced plays a key role in the production and
reception of the message.

7. INTERTEXTUALITY, which alludes to the facts that a) texts are

always related to some preceding or simultaneous discourse,
and b) texts are always linked and grouped in particular text
types or genres (e.g.: descriptive, narrative, etc.) by formal
criteria.
Tischer et al (2000) explain that the first two criteria (cohesion
and coherence) may be considered as text-internal,
whereas the remaining five criteria are text-external.

• Text-internal  TEXT

• Text-external  CONTEXT

Discourse Analisys has given more importance to the external

factors than the “pure” Text Linguistics approaches, but all
approaches within Discourse Analysis view text and context as
the two kinds of information that contribute to the
communicative content of an utterance.
Food for thought: Listen to
Obama’s Victory Speech on the
occasion of his winning the
Elections 2012, and try to explain
how the speech meets the seven
criteria described by de
Beaugrande and Dressler:

http://www.youtube.com/watch?
v=416oyZmugEc
1. COHESION: Is Obama’s speech syntactically cohesive? Do you find instances of, for instance,
ellipsis, reference or conjunction that make the text a cohesive whole?

2. COHERENCE, Does Obama make reference to any elements of the collective knowledge of
Americans? Does he play with the audience’s cognitive structures in order to make his
message meaningful?

3. INTENTIONALITY: What is Obama’s main intention when delivering this speech? Do you
think he accomplishes such an intention? Is it well received by his audience?

4. ACCEPTABILITY: Do you think Obama’s text is well accepted by his audience? Is it appropriate
to the moment in which he is giving it? Are the receptors prepared to receive this
speech?

5. INFORMATIVITY: Is Obama’s speech informative enough, both quantitatively and

qualitatively, to meet the expectations of the audience?

6. SITUATIONALITY: Is the speech given by Obama appropriate to the (social, historical)

situation?

7. INTERTEXTUALITY: Do you find any (implicit or explicit) allusions to other texts or

discourses? What do you think is the type of text/genre of his speech?

Conclusion: After analyzing all these elements, can you conclude that Obama’s speech meets
all the criteria to be defined as a text?
Reflections on some of the definitions of
Discourse Analysis
Schiffrin et al. (2001: 1) note that all the definitions of
discourse and discourse analysis fall into three main
categories:

1. Anything beyond the sentence

2. Language use

3. A broader range of social practice that includes non-

linguistic and non-specific instances of language.
Schiffrin (1994) proposes a balanced approach to
discourse, in which both the formal and the functional
paradigms are integrated. She views discourse as
utterances, i.e. “units of linguistic production (whether
spoken or written) which are inherently contextualized
“(1994:41)

Slembrouck (2005:1) points to the ambiguity of the term

discourse analysis and defines it as the linguistic
analysis of naturally occurring connected speech or
written discourse, or the study of larger linguistic
units, such as conversational exchanges or written
texts. So for him, as for many other discourse analysts,
D.A is concerned with language use in social contexts.
As Van Dijk (2002) or Johnstone (2002) point out, DA is
essentially multidisciplinary and therefore it involves not only
the field of Linguistics, but also:
• Poetics
• Semiotics
• Psychology
• Sociology
• Anthropology
• History
• Communication research
• Political science
• Literary criticism, etc.
SO REMEMBER:

DA is the study of language in use,

including text and context as parts of
discourse.
 BRIEF HISTORY OF TEXT LINGUISTICS
AND DISCOURSE ANALYSIS
In the 20th century, new disciplines emerged within the
field of Linguistics, all of which are interrelated:

Functional Grammars (Functionalism)

Cognitive Linguistics
Sociolinguistics
Pragmatics
Text Lingustics
Discourse Analysis
• All these schools agreed that a good linguistic
description should go beyond the sentence and believe
that if a study is only limited to the syntactic analysis of
sentences, certain meanings and aspects of language
will not be embraced or understood.

• As for the relationship between Text Linguistics and

Discourse Analysis, it may be said that there was a
“progressive integration” of both disciplines, for many
scholars have moved from TL into DA as part of the
natural flow of their beliefs and ideas, as shown by van
Dijk in his biographical article of 2002, where he
explains how his research evolved from Text Grammar
to Critical Discourse Analysis.
Within the category of discourse, then, we
may include not only the “purely”
linguistic content, but also sign
language, dramatization, or ‘bodily
hexis’ (Bordieu, 1990). Thus, it can be
said that discourse is multi-modal, for it
uses more than one semiotic system
and performs several functions at the
same time

One of the discourse analyst’s tasks is to

explain the connection between these
other modes of communication and
language
APPROACHES TO THE PHENOMENON OF DISCOURSE
 Z. Harris (1951,1952) was the first linguist to use
the term discourse analysis and he was a formalist:
he viewed discourse as the next level in a hierarchy
of morphemes, clauses and sentences. But this
view was criticized due to the results of studies
such as Chafe’s (1980, 1987, 1992), where it was
argued that the units used by people in their
speech cannot always be categorized as sentences.
People normally produce units that have a
semantic and intonational closure, but not
necessarily a syntactic one.
 Functionalists , on the other hand, give much
importance to the purposes and functions of
language. Functional analysis includes all uses of
language, for its main focus is the way in which
people use language in order to achieve certain
communicative goals. Discourse is not regarded as
one more of the levels in a hierarchy, but as an all-
embracing concept which includes not only the
propositional content, but also the social, cultural
and contextual contexts.
 Then… what do discourse analysts do?

We can learn a great deal about a discipline by looking at what their practitioners do. Discourse
analysts will explore matters such as:

 Face-to-face conversations
 The language, images, symbols, etc. used in e-mails
 Turn-taking in telephone conversations
 The language of humor
 The use of irony/metaphor for certain aims
 Linguistic politeness
 The discourse of politicians
 Power relations and sexism as manifested in language
 The structure of (written and oral) narratives
 Etc., etc., etc……….
b) THE DATA FOR DISCOURSE ANALYSIS

Before starting, you should know that…

Discourse analysts normally work with some

kind of corpus. A corpus is “a collection of
linguistic data, either written texts or a
transcription of recorded speech, which can
be used as a starting-point of linguistic
description or as a means of verifying
hypotheses about a language” (Crystal,
1997:95)
 DATA COLLECTION
When setting out to analyze discourse, the analyst faces several initial
problems which arise in the form of research questions such as:

What type of discourse am I going to analyze?

This depends on the aims of the researcher, who may focus his study
on spoken or on written language, or may be interested in a particular
kind of discourse or genre (e.g.: political, medical, legal, etc.)

How am I going to collect the data I need, and in case I want to analyze
spoken discourse, how am I going to transcribe and annotate the
data so as to show the features of both text and context as faithfully
as possible?
The criteria for selecting a sample (and therefore tranforming the
material into data) depends on the goals of research. Once I have
collected my data, the approach taken will guide me as to what
procedure to choose (e.g. transcribing spoken discourse, keying texts
in, scanning, downloading material from the internet, etc.)
 TRANSCRIBING THE DATA
When dealing with talk or spoken discourse the researcher turns the spoken
discourse into a document called transcript by means of the process of
transcription. If s/he aims at some degree of objectivity, s/he should try to
use a system of transcription that shows, as faithfully as possible, all the
variables that intervene in the studied phenomenon.

• Thus, a conversation analyst who views talk as interaction would include

not only words in her transcription, but also other aspects such as the
sequential organization of the utterances of the different participants, as
well as interruptions and pauses.

• Some analysts will include information about the text, such as genre, date
and place of publication, etc. Others will include information about the
speakers (sex, age, occupation, etc), or about paralinguistic features such
as pronunciation and intonation patterns, or even laughter.

• But there is no single and accepted way to transcribe speech. Each analyst
focuses on the features that best fit the goal of his/her research
 ETHICS OF DATA COLLECTION

It is an ethical requirement that the

researcher obtain the consent of the
participants, not only to take part in the
study but also to use the data they provide.
Researchers must protect all participants
and observe their legal rights.
 CORPORA FOR DA
Corpus  a collection of linguistic data (written texts or
transcriptions), used to verify hypotheses about a language.
(Crystal, 1997: 95)

Biber et al. (1998:4) highlight the main features of a corpus-

based analysis, namely:

Empiricism (it analyzes the patterns of use in natural

texts)
Utilization of a large and principled collection of
natural texts as the basis for analysis
Use of computers for analysis
Use of both quantitative and qualitative techniques
WHY USE CORPORA?

They facilitate the investigation of language in use

We can use databases of authentic texts thanks to

the aid of corpus linguistics

It allows researchers to analyze patterns of use (e.g.

lexical associations and their distribution across
different registers)
E.g.: the lexical associations for fat, plump, chubby,
overweight
Did you know that…

The first large corpus of English-language

data was transcribed by hand and stored
on index cards which were processed
manually?
This corpus was originally known as the
Survey of English Usage (1960s), and
consisted of a million words included in
200 texts of spoken and written material.
The whole survey was computerized and
is now known as the London Lund Corpus.
 COMPUTER CORPORA AND CONCORDANCE PROGRAMS

Ever since the 1980s, increasingly large

corpora have been compiled
(especially of English). Examples of
Modern corpora are:
 The BNC (British Nacional Corpus)
 The COCA (Corpus of Contemporary
American)
 The ICE (International Corpus of
English)
 The Bank of English

Online Corpora:
 The Shakespeare Online corpus
 The Experimental BNC website
 The Davies Corpus
 CONCORDANCE PROGRAMS
They turn electronic texts WordSmith Tools (collocates and frequency lists)
into databases that can be
searched. Some of them
are:

 Word Cruncher
 TACT
 SARA
 WordSmith Tools
(widely used by
linguists,
lexicographers and
discourse analysts)
Reich’s classification of corpora (1998):
Reich’s taxonomy classifies corpora according to medium, national varieties,
historial variation, geographical/dialectal variation, age, genre, open-
endedness and availability:

a) Medium:
Spoken corpora London-Lund Corpus
Written corpora LOB (Lancaster Oslo/Bergen corpus)
Mixed corpora  BNC

b) National varieties:
British corpora  LOB
American corpora  Brown corpus, COCA
International corpora  International Corpus of English
c) Historial variation:
Diachronic copora  Helsinki Corpus
Synchronic corpora  LOB, Brown Corpus, BNC
Corpora which cover only one stage of language history 
Shakespeare Corpora, Corpus of Old English, Corpus of Middle
English.

d) Geographical/dialectal variation:
Corpus of dialect samples  Scots
Mixed corpora  BNC (which includes samples of speakers from
all over Britain)
e) Age:
Adult English corpora
Child English corpora  CHILDES

f) Genre:
Literary texts corpora
Technical English corpora
Non-fiction corpora  ( e.g.News texts)
Mixed copora covering all genres
g) Open-endedness:
Closed/unalterable corpora  LOB, Brown Corpus
Monitor corpora  Bank of English

h) Availability:
Commercial corpora/Non commercial research corpora
Online corpora/ corpora on ftp servers/ corpora on floppy disks
or CD-ROMs.

This is not an entirely comprehensive taxonomy, for other

variables might be considered depending on the researcher’s
aims.