PREJUDICE IN DISCOURSE

Pragmatics & Beyond

An Interdisciplinary Series of Language Studies

Editors:
Hubert Cuyckens
(Belgian National Science Foundation,
University of Antwerp)
Herman Parret
(Belgian National Science Foundation,
Universities of Louvain and Antwerp)
Jef Verschueren
(Belgian National Science Foundation,
University of Antwerp)

Editorial Address:
Department of Germanic Languages and Literatures
University of Antwerp (UIA)
Universiteitsplein 1
B-2610 Wilrijk
Belgium

Editorial Board:
Norbert Dittmar (Free University of Berlin)
David Holdcroft (University of Leeds)
Jacob Mey (Odense University)
Jerrold M. Sadock (University of Chicago)
Emanuel A. Schegloff (University of California at Los Angeles)
Daniel Vanderveken (University of Quebec at Trois-Rivières)
Teun A. van Dijk (University of Amsterdam)

V:3

Teun A. van Dijk

Prejudice in Discourse
An Analysis of Ethnic Prejudice
in Cognition and Conversation
PREJUDICE IN DISCOURSE

An Analysis of Ethnic Prejudice

in Cognition and Conversation

Teun A. van Dijk

University of Amsterdam

JOHN BENJAMINS PUBLISHING COMPANY

AMSTERDAM/PHILADELPHIA

1984
Library of Congress Cataloging in Publication Data
Dijk, Teun Adrianus van, 1943-
Prejudice in discourse.
(Pragmatics & beyond, ISSN 0166-6258; V:3)
Bibliography, p. 159
1. Discourse analysis. 2. Prejudice and antipathies. 3. Conversation. 4. Minorities. 5.
Ethnic attitudes.
I. Title. II. Series.
P302.D465 1984 401'.41 84-24189
ISBN 90-272-2536-2 (European)
ISBN 0-915027-43-7 (U.S.)
© Copyright 1984 - John Benjamins B.V.
No part of this book may be reproduced in any form, by print, photoprint, microfilm, or
any other means, without written permission from the publisher.
For Philomena
 TABLE OF CONTENTS

PREFACE ix

1. INTRODUCTION 1
1.1. Aims of this study 1
1.2. Theoretical framework 2
1.3. Methods of research 4
1.4. Respondents 6
1.5. Minority groups 5
1.6. Prejudice in other types of discourse 8
1.7. Talk about minorities: An example 10

2. ETHNIC PREJUDICE 13
2.1. Classical approaches 13
2.2. Current research 16
2.3. Toward an integrated framework for the study of ethnic prejudice . 22
2.3.1. The cognitive framework 23
2.3.2. Strategies of ethnic information processing 26
2.3.3. The organization of group schemata 33
2.3.4. The social context 34

3. THE CONTEXTS OF PREJUDICED DISCOURSE 43

3.1. Text and context 43
3.2. Production strategies for prejudiced talk 44
3.3. Social strategies and functions of prejudiced talk 48

4. TOPICS OF DISCOURSE 55
4.1. Dimensions of discourse analysis 55
4.2. Topics of discourse: A theoretical account 55
4.3. Building topics: An example 57
4.4. Topic sequences 61
4.5. Topic change 64
4.6. Contents 66
viii PREJUDICE IN DISCOURSE

4.7. Prejudiced topics 69

4.8. An experimental test 71
4.9. Stereotypes about stereotypes: topoi 73
4.10. Some survey data about ethnic attitudes . . , 74
4.11. Racist discourse: How do majorities talk to minorities? . . . 76

5. STORIES ABOUT MINORITIES 79

5.1. Stories, storytelling, and minorities 79
5.2. Narrative structures 81
5.3. Schemata of stories about minorities 84
5.3.1. The categories of narrative 84
5.3.2. The hierarchical structure of the narrative schema . . 94
5.3.3. Some quantitative evidence 94
5.3.4. An example 96
5.4. Story topics 101

6. ARGUMENTATION 105
6.1. Conversational argumentation 105
6.2. Arguments about ethnic opinions 107

7. SEMANTIC STRATEGIES 115

7.1. The notion of 'strategy' 115
7.2. Semantic strategies in talk about ethnic minorities 116
7.3. Some cognitive implications 128

8. STYLE AND RHETORIC 133

8.1. Strategies of adequate and effective formulation 133
8.2. Some stylistic properties of talk about minorities 134
8.3. The expression of prejudice 136
8.4. Rhetorical operations 139

9. PRAGMATIC AND CONVERSATIONAL STRATEGIES . . . 143

9.1. Speech acts and the structures of opinion interviewing . . . . 143
9.2. Dialogical structures and strategies 147

10. CONCLUSIONS 153

REFERENCES
 PREFACE

This book reports results from the interdisciplinary project 'Prejudice

in Conversations about Ethnic Minorities in the Netherlands', carried out at
the University of Amsterdam. This project has two major aims. First, a
cognitive model is being designed to represent ethnic attitudes in general
and prejudice in particular. Second, an analysis is being made of how people
talk about ethnic minority groups and how such talk expresses their under­
lying prejudices. Empirical data have been gathered in some 120 nondirected
interviews in various neighborhoods of Amsterdam. In the present study we
will focus on the discourse characteristics of prejudiced talk. Only limited
attention will be paid to the social-psychological theory about ethnic
stereotypes and prejudice. In a later study we hope to report more in detail
about this cognitive dimension of the project.
The appearance of this book in the series Pragmatics and Beyond needs
some comment. Although we will also pay attention to pragmatic features
of talk in the narrow sense, that is, to illocutionary functions of utterances
as speech acts, much of our analysis lies 'beyond' this conception of pragmat­
ics. First, also other levels of discourse analysis will be attended to. And
second, we are primarily interested in the relationships between discourse,
on the one hand, and the cognitive and social contexts of language use, on
the other hand. Prejudice and prejudiced talk require an interdisciplinary
account in terms of cognitive models of social attitudes and intergroup con­
flicts as well as in terms of a sociology of communicative interaction and its
context. This means that our study belongs to a broader, empirical approach
to pragmatics, as it was advocated by Charles Morris several decades ago.
Theoretically, however, this research should be located at the boundaries of
discourse analysis, cognitive psychology, social psychology, and microsociol-
ogy.
An important motivation for both our project and the present book is
the realization that ethnic prejudice and racism are a rapidly spreading prob­
lem in our society, especially also in Western European countries. The immi­
gration of large groups of black people from the former colonies and of 'guest
 x PREJUDICE IN DISCOURSE

workers' from the Mediterranean countries has challenged the widespread

myth of racial tolerance in our countries. Within a wider socioeconomic,
cultural, and historical analysis of racism, it has therefore become imperative
to thoroughly study the processes in which racist beliefs and attitudes are
formed and diffused. Besides the mass media, school textbooks, or official
(political, legal) discourse, it is especially informal everyday conversation
among majority members that has contributed to the spreading and accep­
tance of prejudiced attitudes and to possible consequences of such beliefs in
discriminatory interaction with minority members. In this sense, this study
is also intended as a demonstration of the feasibility and necessity of an
applied, critical approach to discourse analysis. To guarantee its readability
for students or researchers from several disciplines as well as for a wider
public of people interested in prejudice, we have tried to keep the theoretical
framework and the terminology as simple as possible. Detailed theoretical
studies will appear as independent articles elsewhere.
We are indebted to several groups of students who assisted us in collect­
ing the interview data for this study, and to the members of the prejudice
project at the University of Amsterdam for many discussions and comments
on earlier versions of parts of this report. We are indebted to Livia Polanyi
for her corrections in the English translations of the original Dutch interview
fragments. The Netherlands Organization for the Advancement of Pure
Research (ZWO) is gratefully acknowledged for its funding of this project.
Special thanks are due to Philomena Essed for her general support and
advice, as well as for numerous discussions about the subtleties of racism as
it is experienced by black minority members. With love and gratitude, there­
fore, this book is dedicated to her.

December 1983, T.A.v.D.

University of Amsterdam
Dept. of General Literary Studies
Section of Discourse Studies
 1. INTRODUCTION

1.1. Aims of this study

In multiethnic societies the different ethnic groups constitute a prom­
inent topic of thought and talk for each other. Especially when new groups
become salient, e.g. by recent immigration, conflicts, or socioeconomic cir­
cumstances, members of the autochtonous majority group will regularly
engage in conversation about such newcomers. Such talk is crucial for the
informal distribution of beliefs and for the expression and social sharing of
attitudes about minority groups. Typically, it is also an important occasion
for the formulation and persuasive diffusion of ethnic prejudice in society.
This study deals with some of the properties of such prejudiced talk among
majority group members. Our examples will be taken from interviews held
in Amsterdam about 'foreigners' in the Netherlands, in particular immigrant
workers from Morocco and Turkey, and black citizens from the former Dutch
colony of Surinam. Yet, our discussion has a wider scope, and hopes to
reveal more general features of racist discourse. In that respect, it may con­
tribute to our insight into talk and communication about minority groups,
and hence into the forms of discrimination and racism in many other 'West­
ern' countries.
Our systematic description of prejudiced discourse is not just an exercise
in applied discourse analysis. Rather, we will focus on those features of
discourse that may be relevant for the expression of ethnic attitudes and for
the diffusion of such attitudes in the community. That is, prejudiced talk is
on the one hand taken and analyzed as a prominent form of social interaction
and of verbal discrimination by majority group members. On the other hand,
it is examined as an observable indication of assumed cognitive representa­
tions of ethnic attitudes and of the strategies for the mental and social uses
of such 'delicate' beliefs. In other words, discourse is both our object and a
method of investigation.
Due to space limitations, however, we will only pay limited attention
to the cognitive and social dimensions of prejudiced discourse, and focus on
the various structures of talk about minorities. Thus, we will investigate
2 PREJUDICE IN DISCOURSE

semantic and pragmatic strategies, style and rhetoric, narrative structures of

stories, argumentation, and other conversational characteristics. In each
case, however, our perspective will be on the specific functions of such struc­
tures in the expression or display of 'underlying' ethnic attitudes, or in the
accomplishment of the (interview) interaction.
Finally, this study should be seen against the background of other
research on ethnic stereotypes, prejudice, discrimination, and racism. Our
cognitive and social psychological perspective on these important social prob­
lems should not obscure the fundamental relevance of sociocultural, histor­
ical, political, or economical factors. And a discourse-analytic approach again
is just one, though rather new, method and object of research within the
area of communication, cognition, and interaction. Yet, by unraveling some
of the details of this everyday, microlevel of prejudice and talk, we hope to
build the bridge between cognitions, on the one hand, and the broader social
dimensions of racism, on the other hand.

1.2. Theoretical framework

Both prejudice and talk are social phenomena that require analysis
within an interdisciplinary framework. A full-fledged theory of prejudiced
discourse, therefore, would be a highly complex undertaking. In this modest
monograph only fragments of such a theory can be spelled out.
A first line of theory formation has been inspired by our own previous
work on discourse (e.g. van Dijk 1972, 1977, 1980, 1981). Although much
of this earlier work does not systematically deal with spoken dialogues in
the social context, it suggests many notions that are also relevant for the
analysis of talk. Conditions on local semantic coherence, the concept of
semantic macrostructure, the analysis of speech act sequences, and so on,
hold both for text and talk. Similarly, our systematic analysis of narrative
structures and their relationships with discourse will appear to be relevant
in the account of stories about minorities.
Secondly, we have drawn suggestions from our earlier work with Walter
Kintsch about the psychology of discourse processing (Kintsch and van Dijk
1978; van Dijk and Kintsch 1983). This cognitive approach provides the
important link between thought and talk, and hence suggests how prejudiced
attitudes and their expression in discourse may be related. At the same time,
this study of ethnic attitudes serves as a possible, social-psychological exten­
sion of the earlier cognitive model of production and understanding.
 INTRODUCTION 3

Since we are dealing with dialogical data, namely interviews, some of

our analyses will draw, thirdly, upon results in conversational analysis, involv­
ing strategic moves in talk, conversational storytelling, turn-taking, sequenc­
ing, and so on (for references, see the following chapters). We will see that
an important feature of talk about minorities is its strategic nature. People
want to make a good impression (and not appear racist), but at the same
time they may want to express their negative opinions, feelings, or experi­
ences regarding ethnic minority groups. These conversational goals may con­
flict, and therefore require strategic resolution, both cognitively and interac-
tionally. At this point, our strategic model of discourse processing is also
relevant (van Dijk and Kintsch 1983).
Fourth, ethnic attitudes in general, and stereotypes and prejudice in
particular, also require more focused theorizing. Our perspective in this case
resembles that of recent research often summarized under the label of 'social
cognition' (Forgas 1981). That is, we view prejudice as a form or as a result
of what may be called 'social information processing', not at the purely indi­
vidual or personal level, but rather as a central property of social members
of groups, on the one hand, and of groups and intergroup relations, on the
other hand (Tajfel 1981, 1982). In this sense, our work is also meant as an
extension of current research in cognitive social psychology about (ethnic
and other) stereotypes, group schemata, and biased information processing
about minority groups (cf. e.g. Hamilton (ed.) 1981.) Whereas much of this
work has an experimental basis, we hope to be able to assess the structures
and processes of prejudice through systematic analysis of natural data.
Another difference with our approach is that in our opinion much of this
work is not cognitive enough, on the one hand, and not social enough, on
the other hand; a well-known predicament of social psychology. For instance,
frequent use is made of several cognitive notions, such as 'schema', 'script',
'categorization', 'prototypes', 'availability', etc., but it is seldom spelled out
in detail what exactly such cognitive representations of social phenomena
(other persons, groups, actions, situations) look like, and what processes are
involved in their actual use in concrete social situations. At this latter point,
a thorough microsociological analysis of interactions and situations is neces­
sary, especially for interethnic encounters. Ethnic prejudice cannot be fully
understood without an explicit account of its functions for observation, action
and interaction in such situations as well as within society at large. The link
between the microlevel and the macrolevel of racism should be established
in this kind of broadly conceived social-psychological approach.
4 PREJUDICE IN DISCOURSE

Finally, prejudiced talk also serves a number of additional communica­

tive and social functions, such as interpersonal persuasion, the diffusion of
social beliefs and opinions in the community, ingroup solidarity, or normali­
zation of attitudes and social precepts for the behavior towards minority
groups. Unfortunately, there is little research about such functions of infor­
mal communication 'through' everyday talk for the spreading of ethnic
attitudes. Yet, there is some relevant work on these social and cognitive
dimensions of communication, from Katz and Lazarsfeld's (1955) classical
book on personal influence until some recent studies on social cognition and
communication collected by Roloff and Berger (1982).
Although this theoretical background is already fairly complex, it should
be emphasized that most of the directions of research mentioned above are
not homogeneous approaches. In the field of social cognition alone, there
are several orientations, such as the more cognitively inspired American
work, as exemplified e.g. in Hamilton (ed.) (1981), and the European
approach as we find it e.g. in Tajfel (1982). Also, it is not difficult to distin­
guish other approaches to ethnic prejudice, e.g. political, ideological, histor­
ical, sociocultural, and so on. In the respective chapters and sections, we
will provide further details and references about theory and research in the
various fields mentioned above.

1.3. Methods of research

It has been remarked above that discourse, for us, is both the object
and method of research. That is, we not only analyze talk for its own sake,
e.g. as part of a theory of some type of discourse, but also to get at the
'underlying' ethnic attitudes of speakers. In that sense, discourse features
serve as data for theoretical inferences about the structures and processes
of prejudice in memory. It is our contention that this kind of natural data
provides insights into the contents and structures of prejudice, and especially
into the functions and other uses of prejudice in the lives of the people that
harbor them, which cannot possibly be revealed through experimental
laboratory work. Of course, such data also exhibit the kind of 'messiness'
that controlled laboratory experiments do not have. Talk is just very much
more complex than responses on scales, choosing between a few options, or
even writing 'free responses' in a laboratory task, and this complexity is also
transferred, as a matter of course, to the analysis of the data. However, the
loss of control with respect to specified outcomes of tested hypotheses is
more than compensated by the richness and the validity of the data from
 INTRODUCTION 5

natural discourse. Many subtleties of ethnic opinions, for instance, would

not be expressed in an experiment. On the other hand, what may be an
experimenter bias in the laboratory may become an interviewer bias in talk.
We will see, however, that this bias is only minimal if respondents can talk
freely. And on the other hand, the interaction with the interviewer reflects
precisely what we want to know, viz. how people talk to others about ethnic
minorities (Schwartz and Jacobs 1979).
At this point, there is another problem, though. Even nondirected inter­
views are of course not the same as spontaneous everyday conversations
(Erickson and Shultz 1982). Even in spontaneous interview talk, people may
tend to follow the strategies of socially desirable answers to questions asked,
especially when minorities are involved (Gaertner 1976). Nonetheless, we
will assume at the same time that also interviews are a form of social interac­
tion and communication, and that many of their properties are sufficiently
close to spontaneous talk to warrant at least partial conclusions about the
nature of everyday conversations about minorities (at least with people we
don't know). The reason we had to content ourselves with interviews is that
it is practically impossible to elicit 'real' conversations about a specific subject.
This would also raise the ethical problem of working with a hidden tape
recorder. Though it is in principle possible to tape a large number of spontane­
ous conversations in various social settings in the hope that the participants
would bring up the subject of 'foreigners', such an approach would not yield
enough data.
Therefore, we have collected a large number of interviews most of which
were held about a purported topic such as 'Life in Amsterdam' and in which
the topic of minorities was often brought up spontaneously by the inter­
viewees, or elso casually introduced by the interviewers.
The relevant portions of the interviews have been transcribed more or
less literally, but only in a few cases in the same kind of detail as is required
for precise conversational analysis. Thus, our data are the transcriptions of
interview fragments, and it goes without saying that also in that case we are
still rather far from what 'actually went on' in the interview. Especially into­
nation, gestures, or other nonverbal features of talk cannot be studied in
that way. Another problem that is relevant for this book is the necessity of
translating the original spoken Dutch into English. For many aspects of
colloquial Dutch, such as the use of particles, this is virtually impossible, so
that the English examples given in our analyses will be only approximations
of the original talk.
6 PREJUDICE IN DISCOURSE

1.4. Respondents
The people we interviewed all belong to the white Dutch majority, and
live in different neighborhoods of Amsterdam. The interviews were held in
three different periods from 1980 through 1983. The first group was con­
ducted in different neighborhoods, that is, both in neighborhoods in which
also a substantial number of minority groups live ('contact' neighborhoods)
and in neighborhoods in which virtually no minorities are present ('noncon-
tact' neighborhoods). The second group of interviews were all held in one
neighborhood in which poor Dutch people live side-by-side with people from
several minority groups. The third set of interviews was held in a rich, non-
contact part of Amsterdam. The underlying rationale for this distribution
was the assumption that ethnic prejudice and especially everyday experiences
regarding ethnic minorities would be different in the contact and the noncon-
tact neighborhoods.
Finally, we did not try to follow the usual rules for the sampling of
respondents, also because many of the contacts had to be established spon­
taneously in public places such as parks, cafés, or shops. Yet, we interviewed
more or less the same number of men as women and tried to speak with
people of different ages. The socioeconomic background of the people was
more or less homogeneous in the respective neighborhoods they lived in.
Although this study focuses on ethnic prejudice, many of the people we
talked with simply cannot be seen as outright racists. In fact, many are very
liberal and tolerant and actually oppose prejudice and racism. In this sense,
it should be stressed that we are not only interested in prejudice, but rather
in more general attitudes about minority groups, whether negative or more
neutral. That is, even if people display tolerance, we want to know how they
do so, since this also is a feature of the ethnic situation.

1.5. Minority groups

Since our examples are drawn from interviews about ethnic minority
groups in the Netherlands, we should briefly specify some of the characteris­
tics of the ethnic situation in that country. Needless to say, a first historical,
cultural, and socioeconomic background for ethnic attitudes and prejudice
is the colonial background of the Netherlands. Both in the East Indies and
the West Indies, Holland, for centuries, had some colonies, and several of
the groups of people that immigrated in the last decades are originally from
those former colonies. A first group of citizens from what is now Indonesia
 INTRODUCTION 7

immigrated after this country became independent in 1948. An important

group of these immigrants constitute the Moluccans, and they have become
one of the major targets for prejudice and discrimination in the Netherlands.
Despite the earlier claims of e.g. Bagley (1973) that — at least until the
seventies — the reaction of the Dutch population to this immigration of
nearly 200,000 people was more or less tolerant, as compared to England,
we witness increasingly negative attitudes also against this group.
Then, in the sixties, the economic prosperity of the Netherlands led to
the employment of large groups of immigrant workers (called 'guest workers'
in Dutch, as in German), first mainly from Italy and Spain, and in the past
decade predominantly from Turkey and Morocco. Whereas the Italians and
Spaniards eventually returned to their home countries, or became more or
less integrated and accepted, the major forms of intolerance in the last years
have been directed against people from Turkey and Morocco. As we will
see in more detail in our interview data, many of the negative attitudes
concern the sociocultural differences and the socioeconomic competition
perceived by members of the majority. Islamitic practices, cooking, the role
of women, and other cultural characteristics of Turks and Moroccans are
in that case interpreted in the same negative way as the assumed competition
for scarce housing, work, and social services. In the seventies, racist parties
have emerged that advocate the return of all (or at least the 'illegal') guest
workers to their home countries, and now have one seat in parliament.
A third major group of immigrants came from the former colony of
Surinam (adjacent to Guyana), which became independent in 1975. In a
short time, more than 100,000 people from that country settled in the Nether­
lands, mainly in the larger cities in the western part of Holland. Whereas
for this group the cultural differences were maybe less marked (most of them
speak Dutch, at least as a second language) than for the immigrant workers
from Turkey and Morocco, the Dutch population was confronted for the
first time with a considerable group of black citizens (African-Surinamese,
often labeled 'creoles', as well as Indian-Surinamese, usually identified as
'Hindustans', besides smaller groups of Chinese and Javanese Surinamese).
Both the Surinamese and the immigrant workers generally suffer from
bad housing, high unemployment, low-paid jobs, and many forms of discrimi­
nation (Bovenkerk 1978). Since the government realized that most 'foreign­
ers' — as they are commonly called — were to stay in the Netherlands
(WRR 1979), policies have been developed to give them special support,
e.g. in education, housing, and social services; but on the whole, these policies
8 PREJUDICE IN DISCOURSE

were often ambiguous. Despite the verbally professed goals of differentation

and cultural autonomy for the various groups, there is at least tacit acknow­
ledgment of the need of integration. And despite strict immigration policies,
many people in the white majority resent the special attention paid to (or
assumed to be paid to) the 700,000 (5%) new citizens by the national and
local authorities.
This concise overview provides some background information about the
ethnic situation in the Netherlands. Not only has a sharp decline in the
economic position of the Netherlands in the eighties brought about a nearly
twenty-percent unemployment rate, but also an increase in openly formu­
lated prejudice, xenophobia, and racism. Whereas 'racism' has long been a
taboo notion to denote the ethnic situation in the Netherlands, the last few
years have brought the recognition that the alleged Dutch tolerance, indeed,
was only a myth. As we will see in more detail in later chapters, recent survey
data have revealed that according to the well-known social distance measures,
more than half of the Dutch population has negative attitudes towards (the
presence of) foreigners, with only slight variations according to region, town,
gender, age, political affiliation, or occupation (Lagendijk 1980). Only about
a quarter of the population appears to be relatively tolerant according to this
kind of interview data. One of the reasons to analyze in depth the kind of
interviews we have held is to gain more, detailed, and especially qualitative
insight into these ethnic attitudes. (See section 4.10 for further data from
survey research about ethnic prejudice in the Netherlands.)

1.6. Prejudice in other types of discourse

Although this book is mainly concerned with the study of prejudice in
conversation, its more general title warrants at least a few remarks about
the expression of ethnic attitudes in other types of discourse. Racism in our
society not only shows itself in everyday talk, but is verbally represented
also in media discourse, in textbooks, political propaganda, laws and regula­
tions, meetings, job interviews, literature and comics, and so on. Only some
of these genres have systematically been examined in order to assess ethnic
stereotypes, ethnocentrism, prejudice, or verbal discrimination. Our project
on prejudice in cognition and conversation, therefore, is part of a larger
framework of studies that aims at the critical analysis of prejudice in other
discourse types, e.g. news reporting in the press (van Dijk 1983a) and second­
ary school textbooks (van Dijk and Spaninks 1981).