Content uploaded by Roy Trevor Williams
Author content
All content in this area was uploaded by Roy Trevor Williams
Content may be subject to copyright.
TEXTS AND DISCOURSES
A framework for the production of meaning
Roy Williams
roy.w.w@ntlworld.com
roy.willliams@port.ac.uk
In: Mass Media for the 90’s. (1993) A.S. de Beer (ed), van Schaik Books, Pretoria, S.A.
pp343-363
In the previous chapter the more traditional quantitative methods of mass
communication research that form the bulk of research in this field, were
discussed. However, qualitative research methods offer very stimulating ideas
and innovative ways of looking at mass media. One such method is discourse
analysis, which is discussed here by way of an introduction, and from a
specific perspective, namely by deliberately challenging the acceptability of
White discourse in South Africa. Four newspapers are analysed in this fashion
(see later in this chapter).
INTRODUCTION
Discourse analysis builds on other approaches to the media. It deals with texts and
with language in minute detail, but within a framework which is a larger unit of
analysis than signs or texts, namely discourse, on which this discussion will focus. My
ideas on discourse analysis owe most to Foucault, critical theory, and various
semioticians including Barthes; to Hallidayian grammar and the forms of discourse
analysis which have emerged from it, broadly called 'critical linguistics', including the
work of Hodge, Kress, Fowler and Fairclough; the writings of the Glasgow Media
Group; and Bruno Latour on the sociology of knowledge. For a more detailed
introduction to linguistic aspects of discourse analysis, consult Fowler (1991).
MEDIA ANALYSIS
Signs, messages, texts, films, television programmes - all of these have been analysed,
evaluated and criticized in media studies. The institutions in which these media
products have been produced, circulated and received have also been examined.
Various kinds of textual analysis have been developed, namely literary criticism (in
various forms), content analysis, and linguistic discourse analysis. In addition, various
schools of thought, from the functionalists through the critical theorists, to
structuralists and post-structuralists, have been applied to the interpretation of media
and media institutions.
Discourse analysis has emerged from an appreciation and a critique of these schools
of thought. What does discourse analysis therefore seek to achieve? To begin with we
can say that it attempts to unify textual analysis and sociological analysis.
BASIC ASSUMPTIONS OF DISCOURSE ANALYSIS
It is possible to analyse texts on their own. But within discourse analysis, this is
problematic. Discourse analysis makes several basic assumptions, which are
numbered and referred to throughout this chapter.
1. Texts are always produced, circulated and received within discourses.
The meaning of texts is obviously dependent on the ways in which signs are arranged
within these texts. But meaning is always assigned within a larger framework or
discourse.
Examples of discourses include: legal, medical, democratic, populist, military,
workerist, and academic discourses. You will notice that many of these overlap.
2. Meaning is normally assigned simultaneously, by the people who produce
the text as well as those who receive it, within more than one discourse.
The next assumption follows from the above two:
3. Meaning is produced within a group of some kind - a discursive
community. Individual variation and creativity obviously also occur, but
usually within the limits of an existing discourse or discourses, and of
discursive communities.
BASIC ELEMENTS OF DISCOURSES
Let us look at a few examples of the manner in which texts are produced within
discourses. One might say that most of the basic ways in which we use signs and
create texts are perfectly straightforward: time and space are self-evident, and they are
surely not the products of some discourse?
Space
Space is defined on maps which have a straightforward reference system, based on a
clear indication of where north is. But why is north at the top? There is no physical
reason why north should not be at the bottom. If we were to turn the globe of the earth
the other way around (upside down), we would not be violating any natural laws. In
terms of the earth's position in space, there is no up or down - we make the choice. It
may not seem obvious that this is part of socialization, in other words, that it is a
social construct rather than a natural fact.
But it does matter. Not in terms of space - because the earth could just as happily be
the other way up. It matters in terms of language and the values that we build into
language. If someone is at the top or ‘on top of’ it is positive. The bottom is always at
least second best. Similarly, going up and going down are loaded with positive and
negative values. One should critically examine the values placed on words and terms
which are used to refer to those in the North and those in the South, and the ways in
which these terms are used to add value to what appear to be objective statements.
This leads to the next assumption regarding discourse analysis:
4. The meaning of words is related to, and even dependent on, the meaning of
other words, and the ways in which they are used.
But to get back to which way up the world should be. Who is the ‘we’ who made the
choice as to which is north and which is south? Certainly not those of us at the
bottom. The answer is straightforward - the people who made the choice were the
people who explored, discovered and defined the world, the people of the North, the
people on top of the world. They decided that they would be at the top and the rest of
us down at the bottom.
The idea of a flat earth, however mistaken it may be, was, in many senses, more
democratic. At the time the ‘top’ of the world was defined, this accurately reflected
the state of world power. The international naval and trading powers of Europe were
on top. Their naval power enabled them to completely dominate world trade, and
therefore the world economy. They had the power to decide which way would be up
and which way would be down.
The world was not defined (only) as a physical object, but also as a text. Not just any
text - a text within a particular discourse of power, which in many senses wrote or
inscribed its discourse all over the world. Even the world - under our feet, was
previously largely undefined, or defined in contradictory ways by different cultures.
There was no single concept of what the world was, that was agreed upon by most, let
alone all people. In that sense therefore, the European colonialists did define the
world, and to a large extent we still live within their definition of it, within their text.
For all practical purposes the world as we now know it did not exist before then.
So when we say that we live in the Southern hemisphere, we are talking within
someone else's text and discourse - the discourse of European conquest. And al-
though, theoretically, we could turn the globe the other way round, this will never
happen. We have produced so many texts within this discourse that to change them all
would be so expensive as to be impossible. We are trapped forever within what we
might have thought were our own texts and our own space.
Time
Time is similarly used as text. It is defined by the position of the earth relative to the
sun. But from which point? Again, the European powers (the British at the time), had
the unchallenged power to make the choice - and they decided that Greenwich, just
outside London, would be the point from which all time would be defined. Greenwich
Mean Time is the primary point of reference. It also defines 0 degrees longitude, so
England is not only at the top of the world, and the point of reference for time, it is
also at the centre of it!
The very idea of time depends, of course, on when it all started. Until barely two
hundred years ago, many people in the West were convinced that the world had begun
about 4 000 BC. We now know that modern man has been around for at least one, if
not two million years; that our ancestors came onto the scene about 10 million years
ago; and that the earth is about four billion years old.
But less than two hundred years ago the first evidence for all this, in the form
of fossilized bones, was not even recognized. There was, as yet, no framework
within which these bones... could be understood. Almost everyone accepted
that the earth had been created as recorded in the Book of Genesis, ... which ...
had even been specified with reassuring exactness by an Irish Archbishop,
James Ussher, who, in the seventeenth century, announced that the earth had
been created on Saturday 3 October 4004 BC at 20:00. Only in the late
eighteenth century-about forty years before the first dinosaurian discoveries -
did scientists begin to work out a more realistic chronology (Man, 1982: 14).
The redefinition of time and space, within the discourse of European conquest, and
later the discourse of natural science, radically changed the way in which people
assigned meaning. Until then, time and space had been defined very differently within
various cultures and religions. Western conquest of the world and of nature radically
changed, codified, and standardized every aspect of time, space, and history - the most
fundamental elements of discourses within which meaning can be produced. The
history of Western conquest has its own ironies. It displaced the fundamental notions
of traditional cultures with its own, in the name of civilization. But this civilization
had its own contradictions, namely those of science and religion. On examining the
discourses of colonialism, one finds many references to the value of truth. But those
truths sometimes contradicted each other.
Even today, some Christians operate within a discourse which contradicts basic
notions of science – evolution and biblical notions of earth history. That should not be
surprising, if one considers assumption #2, namely that meaning is simultaneously
assigned within many discourses. The fact that some of them are quite plainly
contradictory is not necessarily a problem. Thus there is no reason why European
conquest should have found it in any way problematic to civilize people in the name
of truth, which included contradictory truths!
DISCIPLINES OF TEXTS AND BODIES
Discourses, if we want them to include textual and sociological analysis, have to be
defined quite broadly:
5. Discourses are disciplines that order the way in which bodies of texts are
produced, circulated and received; but also within which human and inanimate
bodies are ordered.
This definition is based on Foucault's work (Foucault, 1979). Foucault talks about a
discourse as a discipline - discipline in the sense of a body of knowledge - the
discipline of psychiatry or law for instance. Also, discipline as a practice which orders
bodies - the discourse of psychiatry removes certain people from society, and
incarcerates them in mental institutions. In England, for example, two women were
incarcerated in a hospital in the 1920s because they had illegitimate babies. They were
certified morally insane, and were still there in the 1950s, when two journalists
discovered them. They were possibly ‘mad’ by then. But the discourse of psychiatry
has changed - unmarried mothers are no longer locked up, and husbands can no longer
declare their wives insane.
In 1991 there was a woman in Bophelong Hospital in Mafikeng. She became trapped
within the wrong discourse, with tragic consequences. She had had a baby in the early
1960s, and while in a state of severe depression, she had killed it. She was subjected
to the discourse of psychiatry, within which she was declared insane, and was
committed as a State President's patient. Thirty years later she was still there.
According to the Superintendent of the hospital, a psychiatrist, she appeared to have
recovered from her depression within a few years of the birth (and death) of her baby.
Despite this, she has thus far been incarcerated for thirty years, and the Ministry of
Health in so-called ‘Bophuthatswana’ has no plans to release her.
If she had been dealt with primarily under the discourse of criminal law, she would
probably have been declared temporarily insane, and at most been sentenced to a few
years in prison for manslaughter with extenuating circumstances. She would have
been freed at least twenty-five years earlier. But she was declared insane within the
psychiatric discourse, and she has been incarcerated indefinitely ever since. She is
stuck in the wrong discourse and there is little prospect of her ever getting out of it -
either as a way of classifying her, or as a way of ordering the confinement of her body
- and discourses always do both, although to varying degrees. The way in which one
talks about people - the manner in which one refers to someone - has serious
consequences.
Inscription
Discourses are also about inscribing, or writing, in a wider sense of the word.
Discourses produce texts, such as advertisements, newspapers, and television
programmes. They also inscribe on bodies: the discourse of law inscribes itself on the
necks of people that are hanged. Hanging has recently been abolished in various
countries. This does not mean that the law no longer inscribes itself on the murderers,
it just does so in a different way.
In general, what happens in various fields is that the discourse no longer inscribes
itself on the body of the deviant person, it inscribes itself on the mind of the person
instead - in the case of a murderer, through twenty long years in prison. The result is
that instead of forcibly and often violently excluding deviant people from society,
deviant people are slowly and systematically incorporated into society, but strictly on
society's own terms. You don't fight the enemy, you try to incorporate him or her into
your own discourse. State President F. W. de Klerk' s reform initiative, which started
in 1990, is a fascinating and largely successful example of this approach.
This process of incorporation via the mind is called civilization. It tends to replace the
‘offer you can't refuse’ with something more invidious. South African history is a case
in point. The colonists came with a Bible in one hand and a rifle in the other. They
first eliminated some of the indigenous people (the Khoi and the San), and then
incorporated the rest into Christian civilization. For hundreds of years, whole
continents were incorporated into the (economic) discourses of colonial empires. Now
the discourse has shifted. But the same urge to control and incorporate continues in
the battle for domination, through the trans-national discourses of economics and
marketing. Many of the countries of Africa have discovered that there are no final
victories. The battle simply shifts from the colonial/military discourse to the economic
discourse.
6. Discourses are created and maintained to control the production of texts
and the ordering of bodies. The way in which control is exercised is usually
transformed over time from exclusion and coercion to incorporation and
persuasion - the civilization of power or, the power of civilization.
Apartheid discourse
If the notion of discourse is applied to the South African situation, this might give us
a better understanding of both the notion of discourse and of the historical tasks that
face South Africans in the 1990s, namely to forge for everyone an inclusive civil
discourse for the first time in the country's history.
There is a limit to what can be achieved by means of coercive discourses based on
physical force. South Africa has changed for all sorts of reasons. One of them is that,
in the late 1980s, Afrikaner and English capital realized things had to change if the
economy was to be maintained. Military discourse was not only unsuccessful in
absolute terms (specifically the costly and drawn-out war in the former South-West
Africa (now Namibia) and Southern Angola, and the withdrawal of foreign
investment capital from South Africa, but it was no longer the best option.
The discourse of economic imperialism could possibly be even more powerful and
profitable than military discourse, if only a deal could be struck with the ANC to
‘normalise’ capitalism, preferably in a neo-liberal form. Besides, the discourse of
apartheid was increasingly at odds with the discourse of capitalism, from the time that
demographic factors turned against apartheid in the mid 1980s. The white population
began levelling off, while more skilled Black people, and therefore improved
education for Blacks, became essential for the continued growth of the economy.
Civil and military discourse
In 1990 a period of remarkable transition was initiated in South Africa. South Africa
had always marketed itself as civilized: as a Western Christian Civilization - a society
of civil discourse - but in truth, it has always been a schizoid discourse. Civil
discourse for the whites, and military discourse for the Blacks. The crucial element in
defining civil society, and civil discourse, is the role of the police. In South Africa this
has always been subservient to, and absorbed into, military discourse. Long before the
troops went into the townships, the police were doing active duty in the many
guerrilla wars in Southern Africa, as part of South African destabilization of
independent African states.
The police have always been at the forefront in challenging all Black rights to be in
so-called white South Africa. The dompas, the continual demand: “Waar is jou pas?”
(Where is your ID document?) placed the police solidly within military discourse. It is
only within military discourse, where the boundaries of countries are unstable, and are
continually at stake, that one's presence is constantly challenged. The body has no
established right to be in any particular place. There is nowhere where one can be at
home. This is precisely what Blacks were subjected to in South Africa.
Totalitarian discourse
During the 1970s and 1980s South Africa moved even further away from civil society
and discourse, to what can be called totalitarian discourse - the discourse of the police
state. At least in military discourse your official papers can confirm that you are
legitimately present. As a Black person, the dompas merely confirmed that your
presence was an absence. You were not really here, but really there - there in the 13%
of space that, however much it did (or did not) represent home/land, could never
accommodate you. It was bureaucratic space, not real space. Your space was
constructed in a way that was doubly unreal. You could never be present there, (there
was no work there), but on the other hand, the dompas that always had to accompany
your body ensured that you could never be present here either. You were stateless,
you were put in your place which was actually no place at all.
Military discourse redefines space, and manipulates the boundaries of space. It also
continually challenges your presence. Totalitarian discourse takes it one step further,
and denies any stable presence or state - it administers statelessness. The space that
your body occupies is presumed to be guilty space - culpable until proved otherwise.
Even more destabilizing is the negation not only of bodily presence (although that is
more invidious - you can never get out of it), but of housing. The forced removals.
(and in some cases repeated forced removals) of three million people denied one of
the most fundamental rights - home and land. This became increasingly problematic
because of the media coverage it attracted. So the discourse shifted in the most
significant way - the agency of the discourse changed. For instance, in the case of
terror, one could argue that the Witdoeke at Crossroads in the Cape, (and later the
Rooidoeke of Inkatha), carried out forced removals for the State at a fraction of the
cost, and without anyone being able to pinpoint who was behind it. The State,
operating by proxy and in media darkness, hid its fist - bulldozers are more expensive
and more difficult to hide.
7. Who does and says things within a discourse - the question of agency - is
crucial.
The person identified as the agent can be held responsible. But if there is no agent, no
one can be held responsible, and everyone has to accept that that's just the way ‘it’ is.
The most often quoted example of deleting the agency is the difference between the
newspaper headlines Police shoot strikers and Strikers die in riot. In the first case,
police are clearly the agents, and they may or may not be responsible for the deaths.
In the second case, the strikers are the agents of the action, and they are far more
likely to be blamed. The police, as agents, have been removed, and may even be
entirely omitted in the newspaper report.
Undermining the body and the home as signs of presence is an integral part of
totalitarian discourse. It does not respect the integrity of people. The police in South
Africa enforced totalitarian discourse on Blacks, while simultaneously masquerading
as the bastion of civil discourse in white society.
Military discourse is a zero-sum game. It requires no identification beyond itself, it
has no obligations to public accountability; it acts entirely on its own behalf, and is
answerable only to obedience. It has a licence to kill, and enjoys immunity from
prosecution.
There were many attempts to create the trappings of civil discourse - fairly
successfully until the 1950s in the form of the proceedings of the white Parliament;
much less successfully after that, with efforts which included the charades of
‘homeland independence’, and the tri-cameral parliaments and multiple ministries,
calculated to have cost an additional R6 billion per annum in the early 1990s.
However since 1976, and later in reaction to the tri-cameral farce, Black people took
up the challenge, and responded in kind to the military discourse of the State. If, since
State President F. W. de Klerk's policy statement on 2 February 1990, Black
communities were still operating within military or even totalitarian discourse, whites
had only themselves to blame.
Although the transition from military discourse to civil discourse could be achieved
after 1990 because civil discourse is both more rational and necessary (for resumed
investment, for instance), this will only be assured when it starts to produce the goods
at a very ordinary level - power and effective control by people over their own lives,
and a viable chance of getting on in life within civil discourse.
The minimal condition for civil discourse is unqualified respect for life - and one
wonders if it will ever be attained. We have a long way to go before we can establish
(for the first time ever) a civil discourse which will include all South Africans - in
which all of us will, equally, be citizens.
MARKETING
Much of the bottom line in media is marketing - selling not only commodities, but
also ideas and policies. In the New South Africa, substantial resources are being
redirected into marketing new policies. And this will continue until the political
environment settles down, if it ever does. The discourse of marketing seeks to
incorporate people's behaviour into the corporate strategy of the company,
organization or, for that matter, the State, or whoever else is held responsible for
actions.
People are involved in a variety of discourses, including those of their peer groups.
Successful marketing in discourse terms has to relate to, as well as establish and
maintain, various social discourses, within which its products contribute to and
facilitate desirable discourses other than merely purchasing goods. These can be
called host discourses, within which the market for the product can grow.
Purchasing behaviour can be understood in terms of the various discourses in which
people operate. Some years ago the rising Black market was analysed in terms of
subcultures such as the ‘Mapantsulas’. In the UK a recent marketing analysis on youth
defined not only discourses and subcultures, but also narratives (Euromonitor report
Young Britain 1990). Discourses inevitably include narratives - from the metaphysical
ones about what happens when one dies, to the more mundane ones about what is
likely to happen when children grow up.
The UK study defined groups such as Life's a Party, who want to work for themselves
one day, don't worry about a pension, are not environmentalists, and are generally
anti-Black and anti-gay. Then there are Young Moderates, who will put a lot of effort
into family life and pensions; the Chauvinists who are bigoted and aggressive, who
will work hard and get married, and believe in ID cards for young pub-goers; and the
Safety Seekers, who are nearly 100% middle of the road - they are not convinced that
hard work leads to success, and will not work in a lower paid job now to secure
success later on. The New Moralists are cautious and clean-living, and see hard work
and conformism as the answer to their many insecurities. The Freedom Fighters are
independent idealists, and the Greying Youths are very middle-aged in their attitudes,
believe in hard work, and are already worried. about their pensions (Euromonitor
report Young Britain, 1990). Many of the categories in the UK study were essentially
narrative: pensions, marriage, working hard now for benefits later, smoking,
environmental concern, as well as anti-Black and anti-gay bigotry.
A discourse analysis approach does not, however, maintain that our under-
standing of new information is determined by pre-existing cultural and
political assumptions, but rather that the cultures of any given moment are part
of a social process in which beliefs are produced and contested in the conflict
between groups and classes (Philo, 1990:6-7).
In other words, however much people may have ready-made discourses at hand to use
in the production of meaning, the production of meaning is still something to be
negotiated with the people you are addressing - in discourse - whether personally or in
the mass media. A discourse analysis approach along the Euromonitor lines would
argue that these subcultures are not fixed recipes for interpretation, but rather
discourses within which meaning can be produced, within an interpretative
community or communities.
Shifts in marketing discourse
Marketing discourse, like all other discourses, shifts from time to time. One of the
interesting shifts in marketing discourse, as well as in discourses within the broader
society, is the increasing use of counselling discourse.
The most crucial aspect of discourse is agency. Who acts, who is allowed to act and
who is held responsible for actions, are relevant factors. In psychological counselling
people who are deviant are no longer diagnosed and cured, they are given the
opportunity to share their ideas in a controlled setting, in which they are encouraged
to make up their own minds about how they can function within society - this is the
civilization of mental health care.
The person is allowed to make up her or his mind, but the bottom line is that this
decision invariably has to conform with existing discourses in society, and the person
must end up incorporated within them. On the other hand, a crucial aspect of many
counselling discourses is that they operate in groups and not individually. This is
essential if counselling is not going to be restricted to mere incorporation. In other
words, if it is going to allow for the possibility of change in discourses.
This is because the task of shifting discourses, or even shifting your own position
within a discourse, is difficult if not impossible to undertake on your own. It is
essentially something that involves shifts in a discursive community, not in individual
ideas (if there are such things). (See assumption 3.)
The use of counselling discourse within marketing discourse has, of course, no such
aims. Marketing is plainly in the business of incorporating people's behaviour into
company discourse - and the ultimate example of incorporation is the well-known
claim, ‘What's good for General Motors is good for America’.
Counselling therapy generally assumes that “the effects of social ills can be remedied
on the basis of the hidden potential of the individual" (Fairclough, 1989:225). The
shift to counselling discourse within marketing likewise assumes that the individual
inserts himself or herself into the marketing discourse, using his or her (latent) ability
to make the right choices in a consumer world. The old-fashioned hard-sell which
tries to force people to buy is no longer productive. There has been a subtle but
powerful shift from a discourse in which the company tries to incorporate the
consumer into its discourse, to one in which the company offers the consumer a
spread of opportunities to insert himself or herself into the company's discourse. (A
similar shift occurs in the move from Catholic to Protestant discourse, as well as in
many processes of ‘democratisation’ and ‘secularisation’.)
MARKETING POLITICAL DISCOURSES
To demonstrate the application of discourse analysis to media texts, I will analyse
some aspects of the marketing of political discourses, using examples of front page
coverage of the 1989 general election from The Star, The Citizen, The Weekly Mail,
and New Nation (see appendices 1-4).
The newspapers I have analysed invite the reader to join the struggle (New Nation); to
enter an inclusive South African discourse (The Weekly Mail); to support racist
reform (The Citizen); or to support a slightly less racist discourse of reformist liberal
capitalism (The Star).
Counting
Elections are largely about counting - counting the number of votes. One might expect
counting to be reasonably consistent across the newspapers.
Far from it. Apart from the votes counted for the election, a number of people were
killed and counted as dead on the night of the election. According to the newspapers
under discussion, the number of people killed were: New Nation (25), The Weekly
Mail (23), The Star (more than 4), and The Citizen (none). Three papers agree on the
number injured (about 100), but The Citizen mentions none. Three (different) papers
(The Weekly Mail, The Star, The Citizen) agree on the number of seats won in the
white election, but the New Nation doesn't mention any numbers for any of the tri-
cameral Houses.
The Star (and no other paper) calculates that 75% of the people support reform
candidates, but The Weekly Mail (and no other paper) calculates that a mere 6% of
South Africans supported the Nats - the New Nation does something similar by
dismissing the elections as racist. The Star makes an effort to count the votes and the
swing in key constituencies; and The Citizen lists a few detailed results of specific
seats alphabetically. The other papers do not give figures for any constituencies on the
front pages.
The Star (alone) mentions 130 arrests. Only The Weekly Mail and New Nation
mention a three million stay-away, and The Star is the only paper that mentions (in
small print) any of the results for the “non-white” Houses.
Discourse frames
Discourses have different frames in which meaning is assigned. These frames are
offered to you as a reader, as frames within which you are invited to assign your
meaning. The counting itself - even before we get to the headlines which are more
obvious frames - constructs frames for the various discourses.
The Star's counting frame is narrowly focused on shifts within reform. In fact, it is the
only paper to specifically count the reform vote (75%), so that in case you are
unconvinced that the frame of reform is relevant, the figure of 75% should prove to
you that this is how the vast majority of South Africans (who happen to be white) are
thinking. Within their liberal discourse they do feel obliged to mention the 100
injured, and even the 130 arrests, but they grossly under-report those killed.
The Star and The Citizen reported 24 hours before the New Nation and The Weekly
Mail, which could, perhaps, explain some of these discrepancies. But The Weekly
Mail writes that they received the information early on Thursday from Bishop Tutu,
who released a figure of 23 dead. It is possible, even though it might be debatable,
that the daily press could have waited for Tutu's statement; but they might even have
received and ignored it.
More to the point is the comparison between the resources allocated to counting the
dead in the townships, and the resources allocated to counting the votes in the white
constituencies. Thousands of votes in hundreds of constituencies were counted,
checked and released to the world before Thursday morning. But there was obviously
no consolidated police report on deaths, even though they all occurred within the
comparatively small area of Cape Town, and police always maintain radio contact
with their command posts.
A discourse is not a series of texts and separate contexts. The difference is not in the
texts - one set of figures about voting which is reported, and another set about deaths
which is not. The different discourses are primarily organizational. For the counting
of votes, thousands of people and thousands of Rands were specifically allocated and
managed to ensure that the voting figures were efficiently delivered to the media. In
the State of Emergency, on the other hand, thousands of people and thousands of
Rands (including a whole new State department) were allocated and managed to
ensure that figures on township violence were not available to anyone, least of all the
media (and television coverage was totally banned), which is why Bishop Tutu's
office had to function as a news agency. An examination of the differences in the texts
creates the danger of looking for an explanation in the texts or even in the newsrooms,
whereas the differences are articulated, primarily, at sites far removed from the media
themselves.
The Citizen counts anything it can get quickly and easily from SAPA (the telex news
agency service). It makes no calculations of its own, and is based in a discourse of
'undigested' and 'raw' journalism, in which telex reports are often cut and pasted
almost unchanged onto the newspaper page. The Weekly Mail puts various figures
up-front; in a very conscious working of the figures to create a frame and a headline.
This will be clarified below.
Headlines
The Citizen reports: Small overall majority for NP. This positions the NP (Nationalist
Party) as the party in the centre, threatened on all sides, and it is an implicit appeal for
help. This is thoroughly racist (see below), in that any positioning of the NP
(especially in 1989) in the centre of South African politics is a cynical manipulation
of the facts.
The Star is equally racist, as their headline: Nats mauled on left and right wings could
be substituted for the headline in The Citizen, (and vice versa) without being noticed.
An interesting point here is whether the discourse of The Star and The Citizen should
be called racist or just White. Technically the two are similar: White discourse is
racist by definition. The point is that as a producer of this text I have a choice. I can
choose to call it a White discourse, in which case I am mentioning the facts without
explicitly taking sides - or so it seems. But I would be taking sides. By omitting terms
such as racist, I would be saying that White discourse in the circumstances (taking
into account the White advertising market of these papers, and the White electorate)
was perfectly understandable - and therefore acceptable. By not doing so, and by
choosing to use the term racist, the discourse of this chapter deliberately challenges
the acceptability of White discourse, then and now.
8. Discourse analysis itself is always produced within a discourse; there is no
neutral position from which to analyse discourse.
The New Nation's headline, Election Carnage, is written within the framework of the
Black struggle, but in discourse terms it says a lot more. One of the central
mechanisms of discourse is the placing of agency (assumption 7). The New Nation
implies not only that the election was violent, but also that all the people who
participated in the election were, in an indirect but culpable sense, agents of the
killings, along with the police who are directly culpable. This opens up space for the
discourse not only of the Black struggle, but also that of the armed struggle, waged in
response to the violent discourse of repression.
The Mail's headline does a number of things with very few words and figures, but
many layers of juxtapositioning. It ends up distancing itself from the racist liberalism
of The Star and The Citizen, but it starts by counting tri-cameral votes - it puts them at
the top of the page, in big figures. In this way it enters into the discourse of its more
racist readers, to assure them of the comfort of recognition for their discourse. But it
does so only to lure them on to a different discourse. It is a trap. It invites (marginally
racist) readers to enter a racist discourse, only to lead them on to a fully South African
discourse.
In 1989, South African discourse was necessarily a discourse of conflict, and by
inviting the reader into this South African (as opposed to racist) discourse, The Mail
was trying to create a new discourse of negotiation and, eventually, reconciliation.
Rise of negotiation and in-the-end reconciliation
The Mail's discourse invites the reader to confront and work through the conflict. This
is not the overt (it may, in fact, be a covert) discourse of the liberation struggle. A
stronger probability is that it is a radical reform discourse (liberal reform, if one is less
charitable). It is a pedagogical or educative discourse too - it starts with what you
know, and leads you on to new insights.
The subtlety lies in the 100 Hurt. Without these 100, it would be too confrontational
to be pedagogical, and would fall into the (one-sided) discourse of the struggle. It uses
these hundred to gently introduce the (partly racist) readers to the ‘but’ parts of the
election. If you count the bodies of the voters, the Nats did win. But there are other
bodies to be counted - not only the live White ones, but (first) the Blacks who were
injured, and then, the Black bodies of those who had died. This ironic discourse which
uses the familiar signs of racist discourse to lead the reader into a non-racial
discourse, is also present in the sub-heading. The sub-heading (above the headline)
reads: “Election day: the final results”. It is a parallel invitation through racist
discourse to non-racial discourse. The final results are, simply, the figures at the end
of the counting. But the Mail does not end the counting where you might expect it to
be ended. They count the Nats, then the Conservatives, then the Democrats (in a fair,
liberal discourse of descending order). Then, in a manner similar to the racist liberal
discourse of The Star, they also count the injured, the ‘soft’ casualties of the reform
process. Then they go on counting, and count the dead. In the final analysis, the
election was violent, and the election was also more final for some Black people (who
were killed) than for Whites.
In terms of news values and news discourse this discourse is highly ironic, and
undermines the neat liberal descending order of NP-CP-DP, on two counts: first, the
number of 100 hurt exceeds any of the others, so it should be placed ahead of them.
Secondly, within news discourse bad news always has a higher value than good news,
so the deaths should come first.
The overall result is non-racial discourse, which neatly incorporates and transforms
racist liberal discourse, and which should be a sufficiently subtle bait to entice Star
readers and perhaps even some Citizen readers into its net, and maybe even one or
two New Nation readers, who may feel some rapport with the conflict within the
headline. To be generous, one could call it a discourse of radical pedagogy -
alternatively, reformist non-racialism. The way in which you apply different
discourses to these texts is always as much a statement about yourself as about the
newspapers - describing and ascribing discourses always takes place within your own
discourse - there is no neutral position outside discourses from which you can discuss
other people's discourses. In addition, your own discourse always adds to, and
develops and supports some discourses and displaces others.
Photographs
The photographs reveal another whole layer of discursive strategies and moves.
Because these photographs were taken at the same time, and relate to the same event,
it must have been possible, in principle, for the papers to use any of the photographs.
They could either have bought them from agencies, or they could have sent out
photographers with instructions to get a particular kind of photograph. None of the
photographs are particularly unique. Each photograph that was used can be analysed
in terms of why it was used, and in terms of why other photographs were not used.
9. The absences tell us as much about the discourses as the presences.
THE CITIZEN
The photograph of the (White) De Klerks voting is formally posed. He votes first, and
she dutifully waits to vote afterwards. There are a number of discourses overlapping
each other here, which offer multiple combinations of discursive positions and which
the reader may take up. These include the discourse of male chauvinism, in which he
should vote first. This is overridden by the discourse of political power and authority,
in which his vote counts for more. This is, however, inverted by placing him, the news
value, at the back instead of at the front, possibly because of the intervention of other
discourses, such as fashion (see below).
The discourse of democracy, which makes all citizens equal, is also present: they are
(also) merely a couple - a man and wife casting their (equal) votes - even though in
many ways the votes are also unequal - his vote is a vote that many people will
follow, hers is not necessarily so, and they are also White votes. In addition, the
couple is a symbol of the value of the family, which implicitly calls on a moral
discourse for support.
A sub-element of the discourse of democracy is the discourse of peaceful, legitimate
democracy. The two people in the picture are calm, thoughtful, and rational. The
inscription of the electoral process as peaceful and calm - civilized - legitimises the
democratic discourse. The central value of democracy is that it allows for peaceful
transition from one political dispensation to another. This implicitly contrasts with the
absences on the page - the absence of any mention of the excluded other people -
Blacks. Also, the fact that it is De Klerk himself, who as the White President in 1989,
symbolized the 350 year-old refusal to allow any peaceful transition of power to the
majority. By implication, Blacks are excluded from the election because they are
(supposedly) not committed to peaceful development - probably communists, with no
interest in democracy. This division between peaceful democratic Whites and violent
intimidating Blacks is not questioned in this set of inscriptions. Particularly absent is
the way in which exactly this kind of discourse is an integral part of the structural and
historical violence of apartheid which violated the lives of the Black majority.
The discourse of gendered news values and fashion, in which the clothes that (only)
the woman wears are potentially of interest to the reader, is another possible reason
for putting the man in the background. The discourse of design and layout: if this was
the only photograph available for the front page, it would have to be cropped to
include both De Klerks. There is no other way to show his head and his (voting) hand,
because if she is excluded from the picture it becomes very narrow, and unbalanced.
The overall thrust of the photograph is a discourse of peaceful, home-loving
democrats, acting in a rational thoughtful way as equal and responsible citizens. The
systematic absences on this page, emphasized not as lies but as absences, powerfully
naturalize what is going on here. An election in which a small minority of privileged
people vote within a systematically racist and structurally violent system, and vote
overwhelmingly for its continuation, is presented within layers of highly positive
discourses as totally unproblematic.
Within the transition to a New South Africa in the 1990s, many Whites feel
overwhelmed. This should not be surprising, as the presence of Blacks is new - their
absence has been forcibly maintained and systematically naturalized in everyday life
and in the media for so long. One of the simple yet revealing techniques of discourse
analysis is to ask, what would the opposite be? In this case, it would be a report of
what the whole country thought of its leaders - because that is, after all, what an
election should be. It should be a feedback mechanism to tell the people who govern
the country what the people as a whole think about the way they are governed. In late
1989 South Africa was so divided that it might be difficult to imagine how all that
could fit onto one page.
THE STAR
The picture of Wynand Malan presents him as victor, and perhaps even a potential
saviour - which is the role in which the bourgeois press were systematically casting
Malan at the time - a potential leader to form a new centrist alliance of the right wing
of the Democratic Party and the left wing of the Nationalists.
This is part of the messianic discourse which the bourgeois press tried to establish
from time to time: bringing to the fore figures whom they hoped would break the
political ice (and economic sanctions) with personal charisma. They tried Van Zyl
Slabbert, Malan, Dennis Worrall, and Zach de Beer. Nobody thought of the party
bureaucrat F. W. de Klerk.
The media cannot transform a cardboard figure into a charismatic leader. But they
can, and repeatedly try to, create outstanding figures (literally, within media
discourse) to see whether the human being concerned can successfully step into her or
his media personality. Malan's wife is necessary for the composition of the
photograph, but she plays a marginal role in the picture. It is a clever picture in that it
shows the defeated and dejected Glenn Babb literally being pushed aside. The
dominant discourse in The Star is similar to that of The Citizen; Blacks are
conspicuous by their absence. But this racism is tempered by liberalism - especially in
the subheading at the bottom of the page mentioning four killed and 100 hurt. In terms
of pictures, the appeal of The Star is painfully clear. The large, prominent, very
expensive colour display advert, placed bottom right, a key position in layout terms, is
very revealing.
It is an advertisement to promote tobacco, a damaging and potentially lethal product,
but one which is backed by powerful commercial interests, some of whom have
overlapping interests in media. It is an English brand of cigarettes - that is not only
White, but specifically colonial too, in the case of Mills. Mills is also a brand which
was re-launched after many years, so it appeals to nostalgia. And the advertisement
draws explicitly on elitist English manners and values. The advertisement presumably
works - it must appeal to the readers of The Star, or no-one would pay all that money
to put it there. In terms of layout, the advertisement runs a close second to the
photograph. This is a hybrid discourse - in which reformist and commercial interests
are merged.
The Citizen, in contrast, has removed all display adverts, and shifted the presidential
couple down to the most prominent position in the layout. (Newspapers are read, or
scanned, in a spiral, starting top left, then going down and to the right, across the
bottom, and up to centre left.) There is little else in The Citizen layout that looks
deliberate, but at least the positioning of the photograph fulfils a specific function.
In The Star, the double-line frame of the Mills advertisement is far clearer than that of
the photograph of the Malans, which has no frame at all. And the advertisement is
also more prominent, due to the White space which separates it from the rest of the
text.
When the layouts of the two papers are compared, and if the simple technique of
imagining the opposite is used (imagine no display advert in The Star), it becomes
obvious that both the framing of the photograph in The Star and the headline itself are
crowded out by the Mills advertisement. The headline is, in any event, marginalized
by the messianic photograph, indicating that the dominant discourse of the Star is in
fact (racist) reformist liberalism - a messiah to lead us out of sanctions, and back to a
more profitable relationship with England and the rest of the G7 nations (the West
plus Japan).
Combined with these layers of discourse is the modem discourse of graphics - the
results at a glance, or histograms. Unfortunately, they add no more than a flavour of
modernism. The crucial question, whether the combined opposition could outvote the
Government, is not answered. It is impossible to see clearly (at a glance) whether or
not the Nats could be outvoted. What is worse is that from the histogram it appears
that they may indeed be, whereas this is actually far from true - the opposition total of
72 seats comes nowhere near the Nats' 93 seats. The Star's discourse incorporates
people into unhealthy behaviour in the name of corporate profits, and it naturalizes
racist practices, tempered slightly by some secondary attention to the excluded others,
flavoured with a pretty useless touch of modernism, and optimistic messianism.
NEW NATION
The New Nation photograph shows a potentially menacing armed vehicle, and a
burning barricade. However, the person in the centre seems to be unperturbed by it all.
The discourse of the media Black-out of the Emergency at the time is the reason for
this photograph not being more dramatic. The caption talks of townships erupting,
although this is far from true in the picture. What is explicit, is the extent to which the
SADF/SAP dominates the scene, and inscribes its presence on the public space of the
township. At a more subtle level (probably lost on the readers) the armoured vehicle
has clearly displaced the civil discourse of traffic lights, and has taken over control of
the intersection. The discourse is clearly that of the struggle between unarmed people
and the intruding forces of the SADF/SAP. Everyday life has been suspended. The
New Nation uses green, yellow and Black for its layout, putting in a splash of colour
for the ANC-aligned struggle.
THE WEEKLY MAIL
The Weekly Mail deals with the conflict between the divided parts of South African
society in its photographs, just as it did in the headline. But it has to use two separate
photographs - there is no photograph which, like the headlines, could create a single
text for a South African discourse. The Weekly Mail mixes the genres of photographs
and captions with that of editorial copy. Instead of a caption for the Cape Flats
photograph, they write:
The real mandate for F.W. de Klerk can be found in these figures: On the day
in which six percent of adult South Africans voted for the ruling party, three
million others stayed away from work; over 100 people were injured and 23
killed on election night.
The Cape Flats photograph shows a police officer standing next to a police car, firing
a shot off the frame. The remarkable thing about the picture is that the street scene
continues as normal and the police van does not dominate the scene at all. In contrast
to the New Nation photograph, which seems to be an attempt to show the abnormlity
of township violence, the Weekly Mail makes a very different point - police shootings
are part of everyday life in the townships.
The De Klerk photograph differs significantly from the one used by The Citizen. It is
a picture of De Klerk trying hard, and perhaps too hard (especially as his hand is
positioned on the page to reach out to the words Dead: 23). To add to this, his wife
(also smiling broadly) seems to be pulling his arm back, or at least holding on to it.
The wide-angled lens used in the photograph distorts it, which displaces De Klerk
from the normal - his gesture is artificial, his outstretched hand is artificially enlarged.
The overall perception created is that the townships are normally violated by the SAP,
and De Klerk is artificially trying to convince the (White) electorate of his integrity.
By juxtaposing the two photographs, the Mail questions the validity of De Klerk's
response to the crisis in the townships - where the majority of South Africans live.
CONCLUSION
Discourse analysis assumes a wide framework for the production of texts and
meaning, as well as multiple discourses operating simultaneously. It forces us to see
the text as moves, tactics, strategies, within discourses. This does not imply a
conspiracy theory of media production, it simply means that all texts, from the
simplest conversations to huge media productions, are always produced within
discourses.
Neither the producer nor the receiver of the text (who also produces the text and its
meaning) operates within a single discourse, and neither of them is necessarily aware
of the discourses in which they operate. And they do not necessarily agree with each
other. All the (media) text can do is to offer certain discursive frames in which the
reader is invited to interpret certain readings of the text. The power and sophistication
of discursive frames must not be underestimated. Both the determined media
professional and the determined reader can mobilize powerful discourses in the
struggle for meaning.
REFERENCES
Fairclough,N. 1989. Language and power. London: Longman. Fowler, R. 1991.
Language in the news. London: Routledge. Man, J. 1982. The day of the dinosaur.
London: Bison.
Philo, G. 1991. Seeing and believing. London: Routledge.
FURTHER READING
Anon. 1990+. Discourse and society. (Quarterly). London: Sage.
Cock, J. & Nathan, L. (ed.). 1989. War and society: the militarization of South Africa.
Cape Town: David Philip.
Dreyfus, H. & Rabinow, P. 1982. Michel Foucault. New York: Harvester. Foucault,
M. 1979. Discipline and punish. New York: Vintage.
Hodge, R. & Kress, G. 1988. Social semiotics. Oxford: Blackwell.
Kress, G. & Hodge, R. 1979. Language and control. London: Routledge.
Threadgold, T. 1986. The semiotics of Volosinov, Halliday & Eco. American Journal
of Semiotics. 4:3-4.
Todorov, T. 1985. The conquest of America. New York: Harper.
Van Dijk, T.A. 1985. Structures of news in the press. In: Discourse and
communication. Edited by Van Dijk. Berlin: De Gruyter.
Volosinov, V.N. 1973. Marxism and the philosophy of language. Cambridge, Mass:
Harvard University Press. '.
Williams, R. T. 1983. The impact of informatics on social sign systems. In:
Sociogenesis of Language and Human Conduct. Edited by B. Bain. New York:
Plenum, 353-376.
Williams, R.T. 1989. Discourse & narrative. Context. 2: 5-27.
