analysing talk
and text
a course for the Universitat Autònoma
de Barcelona
Charles Antaki
Loughborough University
lectures   1     2     3     4     5     6     7     8     9     10     11     12  

Lecture/Seminar 10: The Basic Ideas of Discourse Analysis.


Iñiguez, L (2003) Análisis del Discurso (chapter 3). Editorial UOC. Very readable and avoids technical terminology.

Wetherell, M (2001) Themes in Discourse Research: the case of Diana. In Wetherell et al (eds) Discourse Theory and Practice London: Sage Comprehensive and clear.

Discourse Analysis

"Discourse analysis" means many things to many people.

One thing they all agree on is that the analyst's first focus must be on language, and what it does in the world. So far, CA agrees.

The next thing they agree on is that the analyst must 'go beyond' the data itself. The analyst has to interpret by appeal to a theory (e.g. a theory about society, or power, or culture). At this point, DA splits off from CA, which is against interpretation of that kind.

DA then starts to divide into various traditions.

The most general, the one that people tend to refer to if they just say "Discourse Analysis", is a linguistic approach to talk and text that tries to see how the speakers' or authors' choice of words "constructs" a social object. When the data are spoken, this approach is sometimes called "interactional sociolinguistics". Critical DA people explicitly look for the workings of ideology, or power. We'll get to that later.

For the moment, here are two basic DA principles:

1. One of language's functions is to "do things" at the societal level (i.e. above the merely interpersonal)

What things does it do?

It is 'constitutive': some things, at least, are set up and constructed out of language. A good example is The Law: the law in any one society is constituted by all the statutes that Parliament has passed, all the regulations that are written inthe Constitution, and so on. All these are 'just words', but they constitute something very real.

It promotes someone's (or some group's) interests. The Law is a good example again. Discourse analysts want to say that the Law is actually not neutral or impartial (though it claims to be). The language in which it is set up is good for some, and bad for others. The recent debate over how the law treats women's 'murder' of abusive partners is a good example. The question raised by discourse analysts was whether the language of the laws on murder and manslaughter was systematically biassed against women's (alleged) style of reaction to provocation (see "Justice for Women" for a British example of campaigning website on this issue).

2. People use "discourse practices" to do these things.

What are discourse practices?

Discourse analysts (of whatever kind) look for how things are constituted by what they call "discursive practices". If you set out on an investigation into a certain social phenomenon, you will find an identifiable set of things that go together ....

e.g particular words, phrases, terms of reference, metaphors, rhetorical styles, systematisations of knowledge (e.g. rule books, catechisms, manuals, style guides ....)

......which, together, construct that pheonomenon as a certain kind of social object (e.g. 'homosexuality', 'Science' 'Muslims' etcetera).

In each of those cases, the social object is being constructed by the discourse's choice of description, and the associations it implicitly makes.

e.g. the choice between:

Muslim vs Islamic
fundamentalist vs devout

and the association between "Muslim" and ..

terrorism vs insurgency vs freedom vs ....?

Whichever choice you make, and whatever associations you imply, you will help to construct (or 'constitute') a certain social object. For DA (in common with many theories of language in general) the choice of one description over another, and the association of one description with another, is significant. The categories of the world are not ready-made, nor is any use of them neutral.

2. Method: Three examples.

How do you identify a "set of linguistic practices" whch do things? There is no one, single, universally agreed, answer in the literature on DA. In fact many people who do DA resist the notion that there is a simple way of identifying discourses or repertoires. The work of Michael Billig is a specially good example of analysis that explicitly avoids any one 'method'. His chapter in Wetherell et al (2001) is a very good account of that way of working. But there is a range of views, and we can start with those analysts who do base their work on something approaching a systematic, theory-based method.

Linguistic features

There is a tradition among some analysts, especially those who come from linguistics, of basing their work on the systematic grammatical and pragmatic features of the material. A good example of this would be the early work of Gunther Kress (see also his chapter in Wetherell et al, 2001). For example, he pointed out that a very powerful discursive practice in political argument was to change verbs into nouns ("nominalisation"), and to use passive rather than active forms of verbs. So politicians don't say "we are going to privatise the railways". They say "There will be a privatisation of the railways". Kress argues that this choice of grammatical form is a discursive practice. Its effect is to constitutes the transfer of public goods into private hands as 'agentless'. That promotes the interests of certain groups in society.

Identification of "Repertoires"

Other analysts, usually those who come from sociology or other social sciences, are less reliant on linguistics. Many analysts will use their skill and erudition to take us through a text and persuade us how it works to constitute its object by the deployment of what ares sometimes called "interpretative repertoires", or "discourses", or "ideological dilemmas".

Let's look at an example. The one I've chosen to start with is a famous one from the early years of (sociological, rather than linguistic) Discourse Analysis, and is about the discourses of Science. It examines scientists's choice of descriptions of what they and their colleagues do.

Identification of variable 'repertoires'

A 'repertoire' is a more or less coherent way of describing something. It can be a set of words and expressions, perhaps with associated images and so on. If it is a very familiar way of talking about something, we tend not to notice it. For example, many cultures have an apparently familiar way of talking about the perceived world - it is external to the perceiver, 'out there' and factual. But sometimes you will see a person (or a newspaper or other 'author') sometimes use different descriptions to talk about that object. That variabity makes us realise that perhaps the repertoire deserves some attention - just because it is familiar doesn't mean it is neutral. So a useful method of DA is to look for variability: how people's descriptions change.

The observation about variability in repertoires was made by two sociologists, Nigerl Gilbert and Michael Mulkay, in their influential book Opening Pandora's Box (1990), an early example of a systematic discourse analytic study. They identified different 'linguistic repertoires' in the way scientists expressed themselves about their work. In public texts the vocabulary would paint a picture of an empirically knowable real world populated by knowable and secure facts (Gilbert and Mulkay called this the "empiricist repertoire"). In private, however, the scientists' words would change to a "contingent repertoire" which described a shifting world where things could have been otherwise and where facts were humanly constructed. Gilbert and Mulkay found that the contingent repertoire was used especially when things went wrong. If another team of scientists failed to confirm their findings, it was because of 'contingent' things like the other laboratory's poor procedure, or carelessness, or even cheating. It wasn't 'proper science'. Only if things went 'right' would the scientists talk about 'facts' and a regular, predictable universe.

It sounds petty, but it's more than that. The effect of this variability, the Gilbert and Mulkay argued, was to maintain the idea of Science, and defend the principle that there is a knowable objective world. 'Error' is accounted for by human or other failings; 'fact' is arrived at by correct methods. The scientists' discourses of 'contingency' and 'empiricism are defending the very constitution of Science.

Sensitivity to Context

This way of thinking about discourses helped Potter and Wetherell make their classic argument against traditional psychological attitude research (in Discourse and Social Psychology, Jonathan Potter and Margaret Wetherell, 1987). There is a very useful extract of that book in the Wetherell et al Reader (see especially pages 205-207). Potter and Wetherell observed that people can say quite contradictory things (for example, about black people), so the idea that they have one 'attitude' must be wrong. It is better, they said, drawing on the work of Gilbert and Mulkay, to see what each thing they say means in context.

For example, they found a white New Zealander who said these two things about Polynesian immigrants:

What I would li.. rather see is that, sure, bring them ['Polynesian immigrants'] into New Zealand, right, and train them in a skill, and encourage them to go back again


I think that if we encouraged more Polynesians and Maoris to be skilled people, they would want to stay here, they're not as um as uh nomadic as New Zealanders are (Interviewer: Haha) so I think that would be better.

We shall see in the seminar how these two contradictory 'attitudes' make sense in their context. I have put the longer extracts in this appendix.

Next week we shall look at further examples of the methods discourse analysts use.

This week's seminar

Discussion of the idea of 'discourses' on the basis of your reading.

Seminar activity for next week:

Bring in some material from two different kinds of newspaper, having identified some possible 'discourses' or 'repertoires' which might be at work in the text. See if you can also find examples of descriptions which are heavy with implication according to the context.