White South African Elites : a study of incumbents of top positions in the Republic of South Africa 9780702105166, 0702105163

540 53 40MB

English Pages 190 p. [201] Year 1974

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

White South African Elites : a study of incumbents of top positions in the Republic of South Africa
 9780702105166, 0702105163

Citation preview

I

gi t

3

jggg3

f

H.W. van der Merwe

M J. Ashley N.C.J. Chartorr BJ. Huber

~I I

This book is the result of years of teamwork of four social scientists, South African and American: Hzxnztz W. v~ n z z M z z wz, Director of the Centre of Intergroup Studies in Cape Town; M rcHazt. AsHt.zv, educationist at Rhodes Univernty; K~cv C ~ t t r ow, political scientist at Rhodes University, and Bzrrma HtJzzz, sociologist at the University of California,Santa Barbara. This study depended on the cooperation ofnearly one thousand of South Africa's top men and women who devoted a few hours of their valuable time to complete comprehensive questionnaires. The book contains not only details about their socio-economic backgrounds, educational achievements, images of the fitture, views on class and race, and attitudes towards other population groups, but also places them in the w ider context of th e rest of t h e White population. A major finding of the book is the striking contrasts between backgrounds and opinions of English- and Afrikaans-speaking elites; such contrasts are far less marked between the younger members of the Mute population. As it is unlikely that such a study will be undertaken again for many years, this is a u n ique reference work.

ISBN o , o a t

o gr6 3

WHITE S OUTH AFRICAN EI.IT E S A Study of Incumbents of Top Positions

in the Republic of South A&iea

Hendrik %Y. van der Merwe

M. J. Ashley Nancy C.J. Charton Bettina J. Huber

J UTA 4 C O M PANY I T D . E TQWVN

WYNBERG J O ~

ESBU R G

DUR BAN

ISBN 0 7021 0516 3

© H . W. van der Merwe

Printed by Cape R Transvaal Mnters Ltd., Gape Town

P REFAC E The survey of whit e S o uth African elites was conducted during 1968 and 1969, 6rst under the auspices of the Institute for Social and Economic Research at Rhodes University and, subsequently, under those of the Centre for Intergroup studies, formerly known as the A b e Bailey Institute of I n t erracial Studim, at the University of Cape Town. The impetus for this study grew naturally out of my previous r esearch. %York on c o m mumty l e adership p a tterns i n a Canadian and in a South African town proved that the nature of local issues and decisions was in large part moulded by the larger national context. (Van der Merwe, 1962, 1963, 1966, 1967). The subordination of local issues to national concerns is a common feature of contemporary South African society. In view of these observations it was natural for us to turn &om local to national leaders as foci of investigation. This book is the result of a joint elfort of the persons listed as co-authors. I initiated the project and obtained a grant from the Human Sciences Research Council in South Africa. Subsequently M rs Nancy C. J . C h a rton of t h e D e partment of P olitical Studies and Mr M . J. Ashley of the Department of Education at Rhodes University joined me. They assisted in the preparation and administration of the questionnaire and in analysing the data. Miss Bettina J. Huber, then a doctoral candidate in Sociology at Yale University (currently at the University of California, Santa Barbara ), was recruited as a research assistant. She spent one year in South A&ica supervising interviewing in the urban areas. The chapters bear the imprint of their respective authors. Some chapters were written almost entirely by one author while others were joint efforts of more than one person. The authors of the various chapters are indicated in b r ackets in the

table of contents. The style and presentation of material in the various chapters are therefore not uniform, especially since all four authors were highly mobile and spent lengthy periods in different continents while the manuscript was in preparation. The bibliography contains references to several interim reports that have appeared since 1968. In the text only brief reference is made to some of these. Bettina Huber's doctoral dissertation, "Images of the Future Among the KVhite South African Elite", was submitted to Yale University in early 1973. I am very grateful to Nancy Charton and M i chael Ashley for their devoted and pleasant co-operation, their excellent contributions, and their academic stimulation; to Bettina Huber

for the most conscientious way in which she conducted and supervised 6eld work, and for the sophisticated statistical analyses she has made; to Dr Rita Brodie and Mrs Sara Pascall for research assistance; to Mrs Helen Albertyn for secretarial and editorial assistance; to the numerous interviewers who collected information, and to the public who co-operated in the survey. The authorsare indebted to the H.S.R.C. for ha~ing made available a generous grant to conduct this study. The selection of the topic of research, the focus of the investigation and analyses, the presentation of Sndings, the opinions expressed and the conclusions reached are those of the authors and must in no instance be regarded as a reHection of the opinions and conclusions of the Human Sciences Research Council. Cape Town October 1973

Hendrik W. van der Merwe

CONTENTS

Identifying and interviewing the Ehtes (Huaca and VAN DER MKRWK)

1

Definitions ofKlites .

2

Introduction .

Identilying Elite Members and Respondents . The Interview Process Representativeness of the Interview Sample: . Analyses of the elite cadre SVho's Who analyses . Conclusion

9 14 14 15 18

A ProQe of the White South African Klites(HUsza, CHARON and VAv nKR MERwK) 19 Introduction . 19 South Africa's Peoples and Policies 19 The Klites' Location in the White Group 24 Major Characteristics of th e E l ite L anguage Groups . 26 32 Some Causal Patterns The Angiicisation of Afrikaner Elites . Socialisation and Education (Asamv) Socialisation of the Klites: Place of birth . Childhood locale . High school University . Religion Area of Educational specialisation

An Elitist Educations: .

Akikaans and parallel-medium schools English-medium schools . Types of Universities The relative Impact of Ethnic and Struct~ F a ctors on Attitudes towards Segregation and Dissent (HUsza and Vwv nKR MKRwK) Attitudes Towards Legal Segregation . The Impact of Occupational Sector on Attitudes Tolerance of Dissent. Summary

61 61 67 74

76

5 Class Derivations and Perceptions of A&ikaans- and English-Speaking Klites (VAN DER MKRwz ) Introduction . Class Derivations and Present Status: Occupational status

79 79 79 81 83 84 85 85 87 88 89

Intergenerational mobility Education . Income. Class Perceptions Among Klites: Awareness of classes Class divisions Criteria of class Individual Ability and Social Environment 90 Images ofPrestige and Power: 94 Occupational prestige 94 Most powerful an d m os t a d m irable South A&icans 98 Conclusions 101 6 T h e Polity Elites (CHARON) . Delinition. of the Polity Klites in the South The Sig A&ican Political System . Signi6cant Factors in the Socio-Economic Background of the Polity Klites The Field of Experience of the Polity Klites: . Childhood environment of polity elites . Religion Education . Military service Occupational patterns in the legislature Travel abroad

nificance

Interracial contact

Conclusions

104 105 107 108 108 110 115 116 117 118 119

7 Party A l l e g iance a n d Po l i t i ca l P a r t icipation (C HARTON) 121 The South African Political Scene in 1968 121 Political Party Allegiance of Klites: 128 Sector

Home Language Childhood environment Political Participation: Party change Conclusions

128 128 130 131

132 135

8 Some Political Attitudes of the Elites V~

(Can row and

n E R M E aw E.

Definitions of Democracy The Role of the Political Party. Opinions on Nationalised Industry

139 140 142 142

PoliticalTolerance: . The rights of radicals Political intermarriage Evaluation

136 136

145

o f Cer t a i n I n t e r est a n d Pr e s sure

Groups Conclusions 9 I m ages of the Future (HUIER) Preliminary Definitions . The Self-Anchoring Striving Scale, Expectations for the Future. Images of the Future as Causal Variables. A Comparative Perspective: Internal comparisons . International comparisons 10 Whither White South Africa? (Cho w

)

DER MERVIPE

Relations Between Language Groups . Elites and the Masses Beyond the Elites

a n d VAN

151 151

152 156

159 161 161 165 170 170 174 175

Bibliography Index

187

L IST OF T A B L ES Tabk

3 2

7 6 5 4

9 10 11 12 13 14 15

16 17

Page

Number of Elites Approached and Interviewed in Terms of Sector . Klites and white Population by Home Language . S elected Background C h aracteristics b y H o m e Language. Home Language of Klites by Ethnic Background . Childhood Locale by Language Group of Elites . High School Attended According to Language Group. Undergraduate Univemty Attended by Language Group of Elites Father's Religion by Language Group of Klites. Index of Legal Segregation by Home Language Index of L e ga l S egregation b y P o l i tical P arty Preference Index of Legal Segregation by Occupational Sector . Political Party Preference by Occupational Sector Proportion of Respondents Favouring Limited or no Legal Segregationby Home Language and Occupational Sector Proportion of Respondents Favouring Limited or no Legal Segregation by P olitical Preference and Sector. Relationship of Legal Segregation and Dissent . Occupational Status of Respondent's Father, Hi s Own Initial Occupation and His Present Occupation . Occupational Status of Khtes and their Fathers in Terms of Language Group .

Occu pational

5 26 28 34 41

70 71 75

18 Educational Qualifications of K l ites and t heir Fathers in Terms of Language Group

19 Income Distribution of ~ ans a n d English Elites. 20 Views of Language Groups and Political Parties on Social Inequality.

21 Rank Order of Positions by Language Group 22 Language Group of Most Powerful and Most Admir-

92 96

able PeopleAccording to Language Group of Respondent . 23 Political Party P r eference of L a n guage Groups Among Eiites. 129

I'age

? able 2 4 C omparison of P o l i t ical P a rt y S u p p ort A m o ng . Electorate and Elites . 2 5 Political Participation Related to Home Language . 26 I m ages of I n Buence of C ertain I n t erest Groups R elated to Party Preference . . 2 7 G eneral N ature o f H o pes and F e ars b y H o m e Language. . 28 E x pectations for the Future by Home Language . 29 Per Cent of National Samples Expressing Various Hopes . . 30 Per Cent of National Samples Expressing Various Fears. . 3 1 L a dder Ratings by Country: Means and Shifts . .

.

.

.

.

.

.

.

.

.

,

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

1 29 131 146 156 158

1 66 167 169

IDLY'1 IEXIEG AED IJV1 ERVIEAVJVG THE ELITES INTRODUCY1OK

The study of political behaviour in South Africa is still in its infancy. Comparatively little empirical research has been done on such aspects of the political process as election results, party composition, parhamentary procedures and political decision-making. An appropriate preliminary step for such an analysis is a description of the more ~fusible formal structures of the society. This type of formal or structural analysis of the composition of the ruling group in White South Africa would provide a framework for more dynamic analyses of the political process. The project described in the following pages was concerned with what may be termed formal leadership. It i n vestigated those persons occupying f o rmally l egitimised positions of authority in the hope of gaining insight into the South African political structure and process.

In addition, a great deal was discovered about the value system of South African society by the orientations of these leaders. Although such an analysis tells us only part of

analog

the story, it is obvious that the values, career patterns, attitudes,

prides and prejudices of the leaders reveal much about South Africa's prevailing cultural v alues, political ideologies and mobility patterns. While we can better understand society by having knowledge of the elites, it is also obvious that we cannot interpret ehte behaviour and patterns properly unless they are seen in relation to the larger society (see I.ouw, i969: ig). Thu s w e

a t t empted t o d e scribe such relevant c ontextual

features in the following pages wherever they appeared necesWhile the elites do not f orm a d i stinctly separate group, their socio-economic characteristics described in chapter a show the extent to w h ich they are d i stinct from the m asses. Further

analyses in chapters 3, y, y and 8 show how their pohtical views distinguish them from the masses and how class prejudices

I DENTIFYING AND I N T E RVIEw ING T H E E L r r E S

and d i s c r et i o n c o n t ribute t owards separation between upper and lower classes. Our research yielded information about the principles upon which the South A f r i can social systeni allocates elite positions.

We analysed recruitment piucesses among the elites, the social stratification system, views of the class structure, the prestige of certain occupations, and admiration of certain persons. It was hypothesized that preferred social types vary according to language groups and ideological commitments. In chapters y and 5 we analyse contrasting perspectives on race relations and stratification in South A f rica and the contrasting images of power and prestige among Afrikaaiis- and EnglishMost of this book is devoted to describing the results of our in-depth interviews with somewhat over goo members of South Africa's elites. It begins, however, with a discussion of the procedures and problems of data collection and analysis. This chapter contains our definition of positional elites as used in

spiders.

t his study, as well as ou r m e t hod o f i d e n t i fying f i rst a w i d e

cadre, and subsequently a smaller nucleus, of elites within the respective sectors and subsectors. The interview process is aho described briefly, as is an amessment of the degree to which our respondents are representative of the wider elite cadre. In chapter z we o utline the ma jor characteristics of the res-

pondents in relation to th e w ider picture of South Africa's peoples and policies, as well as ms-I1-@is each other. Chapteis S, y and g d eal with v arious social characteristics of elites within the two language groups. We consider their socialisa tion, degree of urbanisation, social class status, and their ethnic and race attitudes. In chapter 6 the polity elites are analysed in some detail. This is followed in chapters p and 8 by a, disolitical participation and policussion of the party tical atutudes of all elites. Finally, chapter 9 deals with the images of the future espoused by the elites interviewed, while chapter Io assems the significance of present trends in South African pohtics, both among elites and masses. While we did not attempt to make specific predictions, we believed that the subjects of our study play a sufficiently important role in South Africa to allow us to utilise our knowledge of them to formu-

allegiance, p

late general trends that m i ght m a terialise in th e f u t ure. DEFINITIONS OF ELITE S

A vagueness about what constitutes "eliteness" is an integral part of much research dealing with various aspects of power structure. The mammoth study of European elites conducted

iDENTlFVLVG AND I N T ERVIEWING T H E

E L IrES

3

from iggg-i965, for example, had no readily acceptable and operational de6nitions of elites(Lerner and Gorden, i969: 38o). Nonethele s, there does seem to be a tendency to think of elites

as somehow "important". As Suzanne Keller ( i963: y ) h a s noted. for example, "Elites are effective and responsible minorities — efFective as regards the performance of activities of interest and concern to others to whom t hese elites are responsible*'. Similarly, Baltzell ( i9gg: i7 g} de6nes ehtes as the most

succ~ful men who are leaders in their respective functional hierarchies (see also Etzioni. i959: 4 j6 ). In addition, elites have been thought of aspeople who are generally revered and respected by the larger community (pratige elites), or as people who have great power (power elites). Writeis like Nadel (i9~6) and C. Wright Milb ( i 95p) have used the concept of prestige ehtes to refer to people who enjoy high status and respect both in the eyes of the society they hve in and in their own eyes. They set the standards and are widely imitated in their lifestyles. Their prestige may derive from their birth, their wealth or their intellect. Power dites, on the other hand, are people who have great political or economic power and who rule their societies. Such elites have been of interest to many social scientists, such as Pareto (i93g), Bottomore (i964) and C. Wright Mills ( i 9g7) to nmne a few. In this study we have limited ourm.ives to a de6nition of elites as incumbents of top positions in society, not because we believed this was the best de6nition, but because it seemed necessary to de6ne a p ositional, or a u thority, elite before isolating South Africa's prestige or power elite. We identified the positional elites through a formal analysis of the social structure, and thereby were able to classify them in terms of structuralcategoric Such a procedure has been recommended by Lasswell and his colleagues(i96z: 6), as well as by Hilda Kuper (i 969: +f) Rnd Wolfga llg ZRpf (1965}~who used suiular approaches in their studies of Indians in South Africa and German elitesrespectively. It is , o f co u i s e, g e nerally r e c ognised t h at t h e p o s i t i onal

leaders (i.e. those occupying leading positions in various social structures) are not always those who exert the most in8uence, who make the major decisions, whose opinions carry the most weight, or who pull the strings. An important task of any researcher interested in leadeiship patterns would be to determine the relationship between the formal and actual leadership. Such an analysis was, unfortunately, beyond our scope in this project

4

I DENTIFYING AND D r r KRVIKw IN G

TH E KLITKS

and remains as a major challenge to political scientists in South Africa. I DENTIFYING KLrI'E NIEMSKRS AND RESPON D E N T S

K lites in s ome societies are u n i ted

i n t o a n e x c l usive e l i t e

group. This has been true especially in pre-industrial societies where power mles were fused in one or a few individuals who lead the nation in government, in religion, and in combat. %Vith the development of functional and structural differentiation in modern society, however, separate elite groups have developed and a single vertical authority hierarchy does not encompass the whole society. Rather, a separate authority hierarchy seems to prevail within each major societal sector or "sphere of activity" (e.g.. economic, political, educational, etc.). This means the society can be viewed as a series of vertical sectors differentiated in terms of function. Each sector is relatively independent of the others and contains its own authority hierarchy or graded structure within which people can climb upwards as they are promoted. As Suzanne Keller (I963: Ba-3) has noted: "Increasingly, no single social stratum is likely to monopolise ace~ to elite positions. The widely accepted model of society resembling a single pyrainid is giving way to one with a number of pyramids, each capped by an elite". Each of these elite groups is very much circumscribed and limited to its own major field(s) of competence. Consequently, positional elites cannot be identif ied without specific reference to that, functional area in w h i c h

they occupy top positions and perform key functions. The failure to do so is a major pitfall in the study of elites.

It has been claimed that the White South African elites form t he type of c l ose-ktut lead ership structure typical o f p r e industrial societies consisting of personal friends acquainted with each other on a f i r st-name basis. Our research has shown that the W h ite South A f r i can ehte structure is pluralistic (i.e.,

various gioups are differentiated, and even separated, from each other, not on]y in terms of their specialimd functions, but also in terms of their values,'. The development of a unified group, or power elite, is hampered in South Africa by ideological conflict on major issues among the elites. This will become clear in the following pages, for much of this book is concerned with the analysis of the ethnic diction of the elites and its implications for views on key political and racial issues. T he pluralistic character of th e e htes emerged early in

our

research. During our i n itial internews it became clear that South Africa was a r elatively complex industrialL~ed society

IDENTIFYING AND I N T ERVlKWING THK

EL 1 TEE

5

with a number of dishnct elite groups whose authority was limited to a specific societal sector. These preliminary interviews consisted of informal conversations with representatives from all walks of life about the formal leadexship pattens prevalent within their occupational spheres. We then wrote to the national offices of labour unions, political parties, churches, newspapers, voluntary organisations, etc. W e a sked fo r i n f o r m ation about their activities, as well as a l ist of t heir to p p o sitions and current incumbents. The response to th~ let t e r s w as very g ood

and only a few organisations refused to supply the information requested. In this way we compiled a l ist of g o3 t

p o s itions and their

z 6o5 incumbents, which we designated the South African elite cadre. From this list we derived a smaller nucleus of elites; that

is, a smaHer "group of p r ime leaders who hold the highest posts in the establishment". (See Janowitz, tg65: g). By this process of selection, we eliminated 678 positions and identified an elite nucleus of z 3g3 positions held by z 927 incumbents (see table t ). All of these incumbents were approached for inten~ews and 925 agreed to participate in the study. Of & e r e st 4to refused to be involved and g9z could not b e contacted, or were unavailable. T AB L E

yYnrnherof EIitesApproached and interoietoed in Terms of Sn:tor

Sector

Persons

Approached '/o Mass Media Education Religion Organisations Professions Monomy Bureaucracy Legislature Judiciary Local Government Total

122 204 124 122 78 446 374 256 66 135

6,3 10,6 6,5 6,4 4,0 23,1 19,4 13,3 3,4 7,0

1 927

100,0

Persons Interviewed 0

/

69 7,5 92 1 0,0 63 6,7 69 7,5 48 5,2 1 90 2 0 , 5 2 12 2 2 , 9 89 9,6 21 23 72 7,8 925 100,0

Positions Interviewed N o/o

85 81 99 9,3 72 6,8 81 7,8 60 5,7 2 40 2 2 , 5 2 40 2 2 , 5 90 8,5 21 19 73 6,9 1 061 100,0

Social scientists have used many d i f f erent c lassi6cations for

analysing social structure and elites. Porter (t96g) distinguished

6

I DEN TIF YI N G A ND n r r E RVIEWING THE E L r f E S

between the economic elite, labour elite, political elite, federal bureaucracy, mass media, the higher learning and the clergy. Mills (I9g6) refers to the Metropolitan goo, the celebrities, the very rich, the chief executives, the corporate rich, the warlords, t he military ascendancy and the political directorate. In t h e m onumental study o f E u r opean elites, referred to a s t h e TEEPS project (The European Elites Panel Survey), oooo interviewswere conducted over a period of I I years. Lerner and Gorden (i969: 9y) started out by interviewing ehtes in the following sectors: governments (high civil service), politics (elected oflicials), business (heads of firms), communications (mass media and intellectuals), church (hierarchy of b ishops and above), miTitary (hierarchy of generals with a f e w strategic colonels), labour unions (heads of n a tional federations and their major components), pressure groups (heads of farmer, wholesaler and retailer federations and their nrajor components). In later years they dropped the church and military sectors, merged labour unions with other pressure groups, and focused on only four sectors: government, politics, business, communicatioiLs.

We initially distinguished ro major sectors in South African society (sce table I ). These societal divisions were rather large and unwieldy, however, and uniform patterns did not always prevail within them, nor was there consistently a single graded structure for advancement. Consequently, the major sectors were further broken down into gg subsectors, listed below: MAss ME DIA

I. South African Broadcasting Corporation s. Press under English control 3. Press under Afrikaans control y. Press organisations EDUCATION

5. 6. 7. 8. 9.

English universities Afrikaans and bilingual univermties University colleges Technical and teacher training colleges Schools

I o. Research institu tcs and Inn sclllns RELIGION

Ir. Afrikaans churches and religious organisations Is. English churches and English-oriented religious organisations

InENTIFYING AND I N T E RYI Ew IN G TH E

E L XTES

NATIONAL O R G ANISATIONS

I3. Cultural and entertainment organisations I g. Health and welfare organisations I g. Community service organisations

I 6. Recreational organisations PROFESSIONS

I y. Professions ECONOMY

I8. Agriculture: unions, co-operatives and control boards I9. Banking and Finance: Banks and banking organisations, discount and acceptance houses, insurance companies, building societies and organisations, stock exchange so. Commercial and industrial organisations 2I. Commercial and industrial companies (including advertising, automobile and oil companies) su. Mining organisations and companies R3. Government corporations sy. Labour

BUREAUCRACY s5. National civil service (including consultants and advi-

sors) s6. Provincial civil setvice sy. Police 28. Defence Force

LEGISLATURE s9. Cabinet 30. Senate 3I. House of Assembly 32. Provincial councillors JUnICIARY

33. Judiciary L OCAL GOY E R N M E N T

3y. City and Divisional Councillors 3g. City and Divisional Professionals In doing so we attempted to have sectoIs that were relatively homogeneous in their field of activity and subsectors that were uniform in terms of organisational structure. Within the sub-

8

I DENTIFYING AND M T E R V I K W ING TH E E I . I T E S

sectors we eventually i d entifiied ke y i n stitutions, organisations or companies and, w i t hi n t h ese, th e to p p o sitions.

The first methodological problem we encountered in choosing respondents involved deciding on the number of interviewees to be taken from each major societal sector. Determination

of this could not be achieved purely in objective terms. In the last analysis, th e r esearchers ha d t o r e l y o n t h e s o m ewhat a rbitrary c r i t erion o f societal i m p o r tance — that is, t h e m o r e important a g i ven sector t o t h e s ociety as a w h o le, th e m ore p ersons were i n cluded f r o m i t . A s ses ments o f " s o c ietal i m portance" were made primarily on th e basis of the researchers'

intimate knowledge of South African society, but wherever pos sible we employed objective criteria to guide our decisions. These criteria included: manpower figures (number of people employed or involved in a given sector); national budget proportions (amount spent o n s ector a n nually), i n c ome statistics (amount derived from sector each year); capital investment within sector, etc. On th e

b asis of such coxmderations relative

proportions were assigned to each sector, thereby determining the interview sample's composition. I n a c t ual p r a ctice the researchers were unable to a dhere to t hese sector proportions as

closely as might have been desirable. In the case of the civil service, for example, it p r oved i m possible to d etermine an ac-

ceptable cut-off point for sample inclusion above the rank of deputy secretary. As a result more civil servants were included

than originally intended. thereby somewhat inflating the proportional representation of the bureaucracy. Unlike the sectois, clusters of subsectors are fairly closely re-

lated to one another and were generally comparable with each other. Consequently, for the purposes of determining sample composition, the proportional representation of subsectors with-

in each sector could be decided upon by using objective criteria. W'hile th e c r i teria used t o a s sign r elative p r o portions varied from sector to sector, the same standards were apphed throughout a g i ven sector or sphere of a c t ivity. I n t h e e conomic sector, for example, proportions were assigned in terms of financial s tatistics such as income, expenditure, profit, abets, an d t u r nover. KVithin the rebgious sector, in contrast, proportions were

based on the membership figures of various major denominations. Determination of the relative number of elite members to be selected from t h e s m aller u n i t s w i t hi n specifiic subsectors required th e use o f c r i t eria s i milar t o t h ose m entioned above. The number of p o tential i n t erviewees chosen from e n terprises

I DKNTIFVING AND LVTKRVIKW ING THE E L f F K S

in the subsectors of banking and finance, for example, depended on a fi r m's assets, that is, the greater the assets, the greater the nu m ber o f r e spondents chosen. I n c o n t r ast, t h e r e l ative nutnbers o f p e rsons included f r o m l o ca l g o v e rnment b o dies

(city councils) depended on the size of the population being served, while the degree of representation from any given newspaper depended primarily on its circulation. Once such proportions had been assigned, it was possible to identify th e actual members of the i nterview sample by select-

ing the incumbents of the highest positions within each organisation. Usually this could be done in quite an objective and straightforward manner by using internal. authority hierarchies. The top positions w;ithin an industrial enterprise, educational institution or civil service department, for example, are clearly defined by the organisation's structure. The lowest ranking position to be included varied considerably and depended on the proportional representation previously assigned to the organisation in question. Thus, while several members of South A f r ica's

largest voluntary bodies were approached for inten~ews, only the presidents of smaller groups were included in the sample. It should be noted that there were cases where internal rank orders

were not officially established (e.g. in city councils, churches, etc.). %%hen such a situation was encountered, the researchers,

with the aid of qualified informants established unofficial authority hierarchies, wluch were valid in the sense that they were generally accepted by members of the organisation in question. City councillors, for example, are officially all of equal status, but it wa s possible to rank t hem i n

t e rms of c o m m i ttee chair-

manships. Thus, such unofficial rank orders within organisations were not merely subjective schemes imposed by the researchers, but were based on standards accepted by the group being studied. 4'hile it was inevitable that some arbitrary deci-

sions had to be made at this stage, as well as during other phases of sample selection, every effort was made to choose potential elite respondents as objectively as possible. T HK INTERV IEW P R O C E S S

The questionnaire we used for ou r eventual i n t erviews with

the South African elites contained fifty-five questions, a number of which had several parts. It was divided into two major sections. Part I asked about socio-economic and more general biographical data (e,g, sex, education, income, religion, language group, residential area, etc.) as well as career patterns (i.e. occupational history and extra-occupational activities) . Part I I

io

IDEVriFYBlo AND nVTERVIEWING THE ELlTES

focused on the respondent's attitudes towards South Africa's future, its social class structure and various political, religious and economic issues. UsuaHy the first part of the questionnaire was completed by the respondent before the interview and required about t w enty t o f o rt y

m i n u tes of hi s t i me. T h e second

part, in most cases, was administered by an interviewer during the course of an hour or so; if the respondent was exceptionaHy efficient the latter part of the questionnaire could be completed in forty minutes, but many interviews ran an hour-and-a-half or longer.

The process of obtaining an interview with a member of the elite generaHy began with a letter containing a brief memorandum explaining the purpose of our research project, a request for a personal interview, and the first part of the questionnaire. Letters were sent to z 9zp elites in the hope that approximately one half of them would agree to participate. This is, in fact, what occurred. After letters had been sent out, one of the res earchers proceeded to the l argest city i n each area to b e covered. A n

a t t e mpt w a s m a d e t o c o n t act e v ery i n c u mbent

of a given position to whom a letter had been sent to ascertain whether or not he was wiHing to participate in the study. Once a respondent agreed to be interviewed, a specific appointment was made by telephone. At the pre-arranged time one of the interviewers met the respondent, usuaHy in his once. The interviewers were mostly local people, hired on a temporary basis. Their task was to coHect Part I o f t h e questionnaire and to complete Part II with the respondent. If Part I had not been completed before the interviewer arrived, it was completed with the respondent during the interview or collected at a later date. After all the potential interviewees residing in, or around, major urban areashad been contacted in the above manner, about three hundred individuals in outlying areas remained unapproached. These people clustered in c ertain subsectors such as agriculture, secondary schools, churches and national organisations. Since interviewing in such outlying areas would have been too expensive, the best that could be done was to mail both parts of the questionnaire to these potential resit. Needless to pondents with a letter asking them to say, this type of approach did not yield very many responses. Even when it was possible to contact elites more directly, about one-third of them (pro) refused to participate in the study. Most refusals were given over the telephone and for generaHy similar reasons. The most common rationale given for refusing to be interviewed was pressure of work. This is

comp lete

XI

not surprising since any member of the elite is, by de6nition, an occupant of a top position whose time is fuIy occupied. Asking him to grant an interview of an hour's duration was, therefore, an imposition. If , i n

a ddition, the elite member in

question had no interest in being studied, or saw little value in any sort of social research (a rather prevalent attitude in South Africa), he was very likely to refuse to participate. Another f actor that o f t en m o v ed i ndividuals not t o g r a nt a n i n -

terview, if not condemn the elite study as a whole, was fear of how the results might be utilized. AIany potential respondents did not believe their questionnaires would be anonymous and were apprehensive lest they fall into the hands of superiors or even the Special Branch of the SA. police (which handles security matters). Other people felt the questionnaire had a "liberalistic slant" and concentrated too heavily on political matters — the latter objection usually came from non-politicians. Some argued that the results could, and probably would, be misused, while others claimed that poliucs were not a valid concern of social research. In view of these objections to the nature of the research project, the constraints on the time and

interest of the elites, and the general resistance to elite studies, it seems fair to conclude that the percentage of people who agreed to be interviewed is quite satisfactory/ Social scientists doing msearch on controvetsial topics can expect to encounter resistance in societies such as South Africa which suffer deep social cleavages. Elite studir3, in particular, are bound to m eet w it h r esistance from b ot h t hose who

oppose the elite and those who protect elite interests. ~ e s uch resistance seldom affected our project directly, it u n doubtedly played a part in the lack of co-operation evidenced by many potential respondents. In view of this it seems appropriate to digress brieRy and consider various aspects of the general resistance to elite studies.

Elite studies have often been elitist-oriented (i.e. students of the elite often happen to view with approval the presencA: of such groups). As a result, students of the elite often encounter resistance on the part of political radicals who suspect such research of being an attempt to justify the superior position of the rulers. This was the experience of Daniel Lerner (an American) when he and his team conducted their wide-ranging "More detailed analyses of the various response rates are contained in the mimeographed report of the study prepared for the Human Sciences Research Council (Van der Merwe, et aL, 1972) and in Settina Huber*s unpublished Ph.D. dissertation (1973: 304-307).

I2

I DKXTIP>DNG AND INTERVIEVANG THK EL I T E S

investigation of European elites; they were haxassed by the political ~t which saw them as agents of the Right (see Lerner and Gorden, Ig6g: 82). Similarly in the South African situation it would have been quite natural for liberals and political radicals to suspect us of believing that SA ute elites have ex-

clu=ive rights and duties to rule the Black masses. In fact, however, informants with liberal political views wele usually willing to co-operate with the research project and were often keen to give as much information as they could. On the other hand, they tended to show concern lest their views, being deviant in

com prom ise

the present pohtical situation, might them in some way. Many of them wanted the promise of complete confidence and the a~rance of the impartiality of the research worker. Some informants wanted t o

k now w h e ther th e i n formation

could under aH cirumstances be kept from the Special Branch of the South African Police.

There are agents of control in South Africa, as in other societies, whose major task it is to protect society against influences disruptive to those in power. One such agency is the government which may promote an unfavourable climate for social research by its unwillingness to promote research and to accept

and use research fmdings. During the past decade, Western social scientists have exp~ thei r c o ncern at the measure of government control ova , a n d i n terference in, their research,

especially in the new nations in Africa. (African Studies Nezesletter, May-June, Ig6g: g; March-April, Ig68: lg ). Resistance to r esearch and i n quiry A p i cally comes from members of the E stablishment who h av e vested interests in maintaining the stattu quo a nd w ho f e ar t h at d i scussion of c ontroversial matters might stimulate dissension and unret. I n

this respect James O. Xlorris' (Ig66: ego', experience in Chile lS Of IntereSt.

Also, Blau reports resistance to his research among Arneric ans questioned about their political opinions, noting that i t

was particularly the case among civil servants. (Blau in Hammond, Ig6g: 3I ). The same resistance was encountered in this study, although on a v er y l i mited scale. At le ~ o n e s enior c ivil servant, however, felt so strongly about i t t hat h e n o t

only withdrew his questionnaire a day after he completed it, b ut launched an a c tive campaign against the project in a n attempt to prevent us from interviewing more civil servants.

One of his major objections was that, even though we promised individual anonymity, we might reveal group characteristics. When asked to explain. he said it would be highly undesir-

Ig

IDENTIC>UWG AND INYERVIEwKNG YHK KL1TKS

able and embarrassing to the government if our fi ndings revealed anti-government a tt itudes among senior ci vi l servants.

The researcher's relationship with the press also tends to be problematic in a society where the press is fiercely partisan. The first problem arises when the press itself, or its personnel, is subjected to an inquiry by social scientists. The second i s a result of th e f ac t t h a t t h e p ress is seldom n eutral on t h e

issues being investigated or in it s attitude to publishing the f indings. Conservative pro-government p apers tend t o b e v e r y

critical of any investigation into controversial topics and tend to publish only findings that support the status quo. Papers in o pposition to the government, on th e o t her h a nd, tend i n

the

opposite direction. The political biases of South African newspapers were clearly reflected in the publicity they gave the first interim report on the elite project. when it was released to the press. The Star of November i, i 9 68, printed the report under the h eading: " N o t

foll owing

M u c h C o n t act of Races," and started with t h e

foHomng words: "A Government-sponsored survey has found a not u n expected lack o f s o cial contact b etween top me n o f

local government and t h e n on-White populations in t h eir cities." The report in Th e Cape Times (July io , i 9 68) was h eaded: " M e n a t

t h e t o p ou t o f t o uch w i t h n o n - W h ite". D i e

Burger gave the report prominence on the leader page and headed it: "Ondersoek openbaar lede in S.A. Plaaslike Besture is Bevoeg". ('"'Investigation reveals that members of SA. Local Government ar e C o i n petent"). Despite this diversity of emphasis, the actual articles in all three papers were factually correct.

It should be pointed out, however, that there are also factors which encourage co-operation of interviewees in an elite study. In a study of corrununity leaders in Grahamstown, South Af r i -

examp le,

ca (see Van der Merwe, i966 ), for only c out of gz respondents refused to complete interview schedules. The explanation for this very favourable response to a controversial questionnaire seems likely to be that inclusion in a selected group of leaders must be flattering to many people. The interview responses indicated a general enjoyment at being included in an elite study and at least one subject openly declared that he felt flattered. ) I e m b ers of t h e i n t erview t eam r eported on

the obvious pleasure some people derived from contemplation of theinselves enjoying this status. Lerner and Gorden (i969: yy) and Slau (ig6y) report the same reactions from elites during their surveys.

Ig

I DENTI F Y I NG AND INTERvIEw lNG THEE Lrr K S

R EPRESENTATIUENESS OF THE INTERVIEW S A M P L E

The aim of this study was to make a series of generalisations about the attitudes of the South African elites as a whole. In view of this, it was advisable to determine how representative the interview group was of the larger elite or elite cadre since not everyone approached agreed to participate. Assessing repr tesentativeness proved t o b e d i K c u h b e c ause background material pertaining to the elites not interviewed was scanty. It was possible, however, to evaluate the elite cadre

in terms of one variable, sector, or that sphere of society in which an individual occupies a position of leadership. Also, t he proportional distribution o f p o sitional r anks could b e analysed within certain subsectors (e.g. economy, civil service, professions, etc.). In addition, it proved possible to investigate selected background characteristics of elite members listed in 8 ho's 8'ho of Southern Africans. All of t hese analyses were carried out by c omparing the variable distributions of th e el ite cadre wit h t h ose of t h e res-

pondent group (i.e., those persons interviewed) and those of the non-respondents (i.e., elite members not interviewed). Irt each case degree of representativeness was assessed in terms of lack of statistical association between the groups being compared. Artrdyses of the elite cadre+ Sufhcient information was available on the elite cadre so that the whole body of non-respondents could be compared with the respondents in terms of sector, subsector and, in m ost cases,

positional rank. Thus, the representativeness of the respondents can be assessed in terms of these variables. A sector and subsector analysis indicates that the interview sample is not different from the elite cadre and, con-

markedly

sequently, th e d i stribution o f r e spondents among t h e v a r ious

societal sectors can be considered fairly representative. There are, however, a few weaknesses that should be noted: the economic sector tends to be under-represented, while the bureaucracy and local government sectors are over-represented. Those sectors proportionally best represented are: education, judiciary, parhament, mam maha, religion, and national organisations.

An analysis of the representativeness of the various ranks within

a u t h o rity h i erarchies revealed tha t i n t h e s e c tors of

aThe analyses dealing with sector and positional rant are ~ more detail by Huber (1973 I 309-318).

d in

I DENTIFYING KID & T KRVIEWING THE ELIT E S

l5

economy and mass media, directors tend to b e under-represented while managers are over-reptesented. This seems to be due to the fact that directors were difficult to contact through company offices, a conclusion that is buttressed by the fact that managing directors tend to be over-represented among the interviewees.

In addition to the economy and mass media, it was possible to analyse the positional distributions of the civil service, local government, national organisations, and professions. Within each of these groups the respondents are highly representative of the larger elite in that sector or subsector. With the exception of professions, such slight trends towards lack of representativeness as there are, are similar to those observed for the economy and mass media. That is, "higher" positions (e.g., director, elected oScial, cabinet minister, etc.) tend to be underrepresented among the respondents, while the " l ower" managerial and appointive positions are over-represented. This implies that the interview sample may be shghtly biased in favour of a certain type of elite, namely the managerial types responsible for the day-today decisions necessary to keep a business enterprise, a governtnent department, or voluntary organisation running smoothly. 8 ho's N%o analyses+ While the analyses just summarised served to give some indication of the interview sample's representativeness, it seemed desirable to undertake further comparisons. Initially, the diKculty with doing so lay in a subsample of the elite cadre (containing both respondents and non-respondents) for which more detailed biographical information was available. The source for such a sample was found in the reference book, N%o's i@ho of Southern Africa, (the Ig68 edition was used). This book contains a good deal of background material for the persons listed, and since it included about one-third of the elite cadre, its use allowed further comparisons of respondents and non-respondents. In order for the subsample drawn from 8%o's 8'ho to reveal that the interviewees are representative of the larger elite, two things had to be established:

iden tifying

(I) that the 8'ho's 8'ho sample (i.e., that portion of the elite cadre included in the book) is representative of the elite as a whole; and +The Wi'so's Who analysis is dealt with more extensively by Huber

(1973; 318-326).

16

IDENTIFYING AND INTERV I E W ING THE EI ITE S

(a) that there are no significant differences between the respondents and non-respondents listed in Who's Wl'Io in terms of t h e b ackground characteristics selected for analysis. The first point was substantiated by comparing the Who' s WIIO sample with the elite cadre in terms of two variables: interview status and sector. This analysis showed that in terms of rela-

tive proportions of respondents and non-respondents, the Who' s Who sample and the elite cadre are almost identical; th e inter-

viewees make up approximately one-third of each group. Consequently, the WIIo's WIIo sample is highly representative of the elites in terms of i n t erview status. Th e sector analysis also reveals representativeness, though of a m ore moderate sort. Vfhile the overall distributions of the tw o g r o ups being compared do

not differ greatly, there is a tendency for the economic elites to be over-represented in the Who's Who sample, while the sectors of national organisations and p are somewhat underrepresented. This bias results from the fact that N%o's Who of Southern A jrica is geared to the business community, and consequently includes an unusually large number of persons occupied in the country's economic enterprises. It should be noted

rofessison

that the WIIos' WIIo sample and the elite cadre were also comp ared in t e rm s o f s ubsector distributions, but n o n e w t r e n ds

emerged. The results of these analyses indicate that the Who' s N%0 sample is relatively representative of the South African elite cadre identified for the purposes of this inv~igation. Consequently, trends found within the former group were aLm encountered in the latter. For the elite members listed in W h o's WII0, information was

available on th e foBommg characteristics: sector, sex, age, marital status, level of education, area of residence, and country of birth . I n

o r der to assess the representativeness of the in-

ter~iew sample in terms of these variables, the respondents and non-respondents listed in W h o' I

W I I o w ere compared with one

a nother. It was found that the two groups do not di e r t o o greatly in terms of their sector distributions. There is a. marked tendency, however. for non-respondents to be di~ r oportionately clustered in the economy, wh ile the sectors of professions

and religion contain unusually large groups of interviewed elite members. It i s o f i n t e r est to n o t e t h a t t h i s b i as i n t h e r e -

spondents' distribution is essentially the same as that identified in the earlier analysis of the interview sample as a whole. In addition, this result lends further support to the conclusion that

I DENTIFYING AND IWTERVIEw IN G TH E E l . rr ES

the 8'ho'I 8 ' h o sample is representative of the elite cadre.

The social characteristics for which data were available reveal lesser differences than the sector analysis did. Only one of the comparisons involving these variables revealed a w eakness

worthy of mention. The distributions of respondents and nonrespondents in terms of geographical area are not as similar as they might be. That is, respondents are more likely to live in the Northern Transvaal and Eastern Cape than non-respondents, while the latter are disproportionately clustered in the Orange Free State and, to a lesser extent, in the Western Cape. I n addition to geographical distribution, it might also be noted that t he respondents t en d t o b e s o m e w ha t y o u n gel a n t i b e t t e r e ducated than the non-respondents. In spite of these minor i n -

c onsistencies, it can be concluded that the Who'I W h o r e spondents and non-respondents do not differ greatly from one another in terms of the six background characteristics analysed. In view of the earlier findings that the who'I 8%o sample is r epresentative of t h e e l it e cadre, t his conclusion may a lso b e

applied to the South African ehtes as a whole. This implies that the elites interviewed are representative of the alite cadre in terms of age, sex, marital status, educational level, area of resi-

dence, and country of birth. While the FYho'I f4~ho analysis showed representativen~ on the social characteristics analysed, it di d r eveal some weaknes~ in terms of the sector distribution. Of possible relevance to any conclusions the intervie~ data yielded is the fact that the economic sector is generally under-represented, while the political sectors of bureaucracy and local government tend to be overr epresented. In v i e w o f t h e S outh A f r i can c o ntext, i t i s c o n -

ceivable that, this configuration might lead to a slight conservative bias in the data. In addition, the positional composition of the sample might abo have an impact on the interview results. While the respondent group is largely representative in terms of positional rank, there is a tendency for the "upper crust" (e.g., directols, cabinet m i nisters, elected of5ciajs, etc. ) to be underrepresented. Also, the economic segment of t h e respondent group is not representative in terms of positional rank, thereby making it the least well-represented s~ e n t of the elite sample. A s long as t h e l i m i t a tions on r e p resentativeness just m e n tioned are borne in mi nd, which we have tried to do during our

data analysis, it can be assumed that the respondents' opinions described in the following pages reflect. those of the larger elite cadre, The trends we found within the interview sample can, by

I 8

I DE NTI FY I N G AND INTERVIEWING THE ELITES

and large, be considered indicative of characteristics pertaining to the South African elite as a whole.+ CONCLUSION

This project is in the nature of an exploratory probe into a field of social research which has enjoyed little attention in South Africa. While the research is concerned with "dites", we have tried to refrain from making it "elitist" in intention or in tone. In order to obviate dMiculties of definition we chose to utilise the readily available authority framework. Even this presented difliculties in identifying a sample. But it was possible to sdect one fairly objectively on this basis. We admit that the sample, in giving representation to those in positions of authority, will tend to overemphasise attitudes which are resistant to social change. We admit, too, that the many restraints operating in the South African situation, including our own position in the society, will tend to influence the nature of the responses we received and our interpretations of them. In this situation as social scientists we can only faithfully try to recognise the bias to which our observations might be subject,. This we have attempted to do. We consider that in spite of all these diKculties and disadvantages our work has validity, and can be utiTised to describe and interpret the South African social and political scene and the role of the elites within society.

«Our claim for making inferences about all South African elite- receives additional indirect stqiport from the 6ndings of The European Elites Panel Survey (TEEPS) which included substantial tests for reliabiTity and made several inferences of i mportance for elite studies in general, On the basis of an analysis of the responses they recrived, the researchers concluded that "elite surveys require many fewer respondents than previously considered necessary to achieve a reliable measure of ehte opinion". (Lerner and Gorden, 1969: 9$). Our 925 respondents, therefore, are a large enough group to represent adequately the opinions and orientations of the larger elite cadre comprising about 2600 persons.

CHAPTER II

A MOEILE OI' TKE HHITE SOU1H AEMCAE EI.IZES INTRODUCTION

To understand the significance of ehte attitudes on a wide variety of subjects, it is necessary to appreciate the larger context within which they operate. This chapter, therefore, tries to set the stage for subsequent discussion by locating the elites we studied within the larger South African context. We begin by briefly sketching the major characteristics of South Africa's four racial groups: the Africans, the Coloureds, the Asians and the Whites. We then analyse the elites within this broader national context by comparing them with the larger White group of which they are a part. The comparisons reveal that although the elites are significantly different from the mass of people in terms of a

n u m ber o f s o cial an d economic f a ctoi3, these dif-

ferences do not lead to a distinct cleavage. It appears that the gap between elites and masses in South Africa is not great enough to have a decisive inRuence on the social, political and economic relations or policies of the country. Once the elites have been anchored in their larger social setting, we turn our attention to another factor of importance for understanding elite attitudes: t h e i n ternal g r oupings and cleavages that differentiate various segments of the interviewed group. Much of this chapter is devoted to a lengthy discumon of the socio-economic characteristics of the ehtes studied. Since one of our principal findings was that elite attitudes were distinguished most sharply in terms of ethnically linked attributes, the socio-economic discussion is articulated in terms of language differences. The chapter concludes with a look at how ethnic allegiances have changed with time and we speculate brieRy on the significance of such change. SOUTH AFRICA S PEOPLES AND POLICIES

The population of the Republic of South Africa is both racially and ethnically diverse. The Whites, numbering 8 75I ga8 or I7,5 per cent of the total population in Ig7o, constitute the ruling group and hold both economic and political power. The

20

A PRO FI L E OF THE WH ITE SOUTH AFRICAN ELITES

vast majority are urban — 86,56 per cent. The White group is ethnically divided into two major cultural groups, the majority Afrikaans group and the English-speakers. They govern country through structures which eInbody the classic principles of representative democracy as far as the Whites are concerned.

There is a parliament consisting of a State President, a Senate, and a House of Assembly. Parliament is the "sovereign legis l ative in and over the Republic". (Constitution Act, I 9 6 I .) Judicial authority is vested in the Supreme Court of South Africa with Appellate and Provincial divL~ions. Judges are appointed by the State President on the recommendation of the Cabm They enjoy security of tenure and can be removed from office only on an address from the Senate and the Assembly for mis behaviour o r

i n c a p acity. T h e c o u rt s a r e n o t e m p o w ered

judge the validity of an Act of parliament except one repealing or amending the entrenched clause which protects the status of Enghsh and Afrikaans as official languages. With this one excep tion then the constitution is fully flexible, and parliament is supfelne.

The State President is elected for a period of seven years by an electoral college consisting of members of the House of AsM;mbly and the Senate. He is merely a ceremonial head of state and has little or no p ower or d iscretion to act on his own i n i t i ative, I n

Ig68, the House of Assembly consisted of I66 member, elected by umversal White suffrage. In addition there were four members representing the Coloured population. This separate represenmtion has now fallen away. The Senate consisted of gy members, of whom 43 were elected

by the members of parliament and members of Provincial Coun cils for each province and South West Africa. Ten members wete nominated, two from each province and South West Africa, and there was one additional member nominated to represent the interests of the Coloured people. The Senate is subordinate to the House of Assembly, having no effective power over the raising and spending of money, and delaying power only over other types of legislation. The most powerful body in the South African constitutional framework is the Executive Committee, or Cabinet. This consists

of eighteen ministers who are members of the House of Assembly or the Senate, appointed by the State President on the recom mendation of the leader of the majority party in the House of Assembly. They hold office until defeated by a motion of no con Mence or until they resign voluntarily or are defeated in a gene ral election.

A PROFILE OF THE WVHITE SOUTH AFRICAN ELITES

This brief outline of th e structure of th e c entral governtnent reveals several ~ can t fact s : T h e ce n t r a l d e c ision-mak-

mg agency is all White and it represents, with the exception of the one nominated Senator, only W h ite int e n t s . T h e H o use of

~m b l y h olds the supreme legislative power and it is wholly White and response)le to a W h ite electorate. Supreme executive p ower is held by a n E x e cutive or C a binet, also wholly W h i t e

and representing only White interests, and responsible to the Legislative Assembly. AfrI~ co n sti« t « e ma j o rity group numberingI $ og7 gyp or 7o,a per cent of the population. Although the Afric~ s t em f rom a s u m lar c u lture b ased on a s u bsIst c e e c on om , t h

are ethnically divided into nine major tribal groups +~k;ng different languages, not always Inutually intelligible, and havmg diferent social customs. They have been allocated I3 per cent of the land, mainly r ur al, which they hold fo r th e most part in

communal tenure. One-third are permanently domiciled in the urban areas of White South Africa, A further aI per cent live the rural areas of White South Africa. Territorial segregation, which is the aim of t h e present goverrnnent, thus reInains more of a myth than an attainable ideal. The 46 per cent which

are permanently domiciled in the homelands are poor, and in Inany areas there is distressing evidence of the pressure of exist-

ing population on resources, Economically the African population occupies the lowest rung in th e social hierarchy for although they have been drawn into the Inodern economic system of the Whites, it is mainly as poorly paid wage-labourers. In I968 it was estimated that the per capita purchasing power of the African was R8z per annum in comparison with RI6g for C oloureds, M 8 3

f o r A s i ans and R I I4 I

for W h i t e s . (G.

Leistner in N. Rhoodie (ed.) I972: z84). There are minor concessions to the plural nature of South Afri-

can society within the political structures. The Bantu Authorities A ct of I 9 5 I , t h e Pr omotion of B antu Self G overnment Act o f I9g9 and the Transkei Constitution Act of I 9 6 3 , among others,

set up a hierarchy of local government structures designed to articulate the i n terests of

A f r i cans. T hese structures amount

today to "blue prints" for " embryonic states". (Cf. Worrall, I97 I ). Only in one region, in the Transkei, has significant progress been made towards local self government, and even then the Transkeian legislature is completely subordinate to the central government, structures described above.

The Coloured people,numbering 2 oI8 4g3 or 9,4 per cent of the population, occupy a slightly higher rung in the econo-

22

A PRO F I LE OF THE WH ITE SouTH AFRICAN ELITES

mic and social structure. The great majority (7g,p7 per cent) are urban and are almost wholly dependent on the modern economy. Many are skilled and setni-skilled workers which accounts for their slightly higher purchasing power in comparison to the Africans. The Coloureds do not have a homogeneous culture and there are sub-groups which differ frotn each other in language, religion, racial derivation and life style. In I968 the Coloureds were disenfranchised with respect to the central parliament and their communal representation in parliament was abolished. For them there is now the Coloured Persons Representative Council with go members elected by univeI3al Coloured franchise and ao nominated members. At present this body has only minor legislative functions and no taxing power; it is wholly subordinate to the White legislature. Unlike the African political structures it is c ommunaUy based. The African structures are all territorially and communally based; the territorial centre is the tribal "homeland" or "heartland", and all l egislatures enjoy mtra-temtorial jurisdiction over the members of their respective groups in some respects thus making them communal as well as territorial authorities. The Asians,numbering 62o4g6 or u,9 per cent, occupy a middle position in the South African social and economic hierarchy with a slightly higher economic ~tus than the Coloureds. The Inajority are urban (86,77 per cent). All are dependent on the modern sector of the economy. Many are skilled and semiskilled workers and a sizeable number are successful entrepreneurs. The Asians, too, are not homogeneous, being divided into two major groups which differ in religious a%1iation. For the Asians there is the South African Indian Council, an appointed body with advisory powers only. The development here is the most rudimentary of all and Worrall (I97I) remarks that it is difficult to judge the potential of this Council in its present form. Although South Africa is a plural society its effective governmental structures are purely unitary. Concessions made to the subordinate racial groups with regard to the separate articulation of interests hold promise for the future but do not yield effective power at the moment. Not only is the effective power held by the white minority, it is also centralised from a territorial point of view. There are White local government bodies in the Repubhc but aH are strictly subordinate to the central structures. South Africa is thus a highly centralised unitary state. All important effective decisions are taken within the central structures, manned solely by a section of the White minority.

A PROFILE OF THE wM T E SO UTH AFtuCAN ELITES

2$

Another important aspect of the South African political structure which dcscfvcs attcntlon ls thc cxtcnt Qf cxccutlvc domm

ance in the governmental fralnework. "While it is true that developments in parliamentary systems in the past half century have tended to exalt the executive at the cost of the legislative, this tendency has been carried to its furthest point in the South African governmental p stem". (Hanis i n W o r rall, l9 y )l . Harris goes on to define the South African political formula as the "logic of' cabinet concentration". Judicial and legislative powers have been delegated to th e executive in increasing measure, sometimes in answer to the increasing governmental involvement in the hfe of society, which is characteristic of most industrialised states, and sometimes in answer to the security needs of the White population. The government has ahvays had powers to legislate by decree in the African areas. And outside these areas their powers have been groi~mg in response to a need for regulation created both by the prevailing ideology of separate development, and by the crisis situation induced by the growth of militant African nationalism in the l9gos and l96os. The rule of law has been abrogated in many itL~nces, and Cabinet Ministers, government o%cials and

police o%cers have been given legislative and judicial powers which affect the lives of all South Africans. "In the government's determination to preserve 4%ute supremacy, which it equates with the safety of the state, the South African authorities have replaced the rule of law with the rule of Cabinet Ministers, public O%cials and police o%cers". (Tholnpson, l 966). Profer Ben Roux (in Worrall, l97l) remarks that the results of this increasing governmental involvement in the life of society has been a curtailment of the judiciary's powers, decision-making by commission, complete pragmatism in inter-governmental relations, increasing participation by top o%cials in the policymaking process, the impossibility of p r eserving a p ersonal touch in the execution of public business, and the growing difficulty of ensuring public accountability on the part of some executive agencies. Relations among the various race and ethnic groups in South Africa have ahvays been marked by patterns of both integration and segregation. Initially the segregation between Slack and White was simply an unformulated "way of life", a set of tacit assumptions among the voting gloup. Since the coming to power of the National Party, holvever, the "way of life" has been formahsed into an elaborate ideology which constitutes a rationalisation for separation.

24

A PROFILE OF THE ~

SOU TH A F R I CAN ELITES

Historically the development of the economy and the demands of the new urban industrial set-up tended to bring groups into closer contact; race and ethnic prejudice, group interests, cultural and historical diHerences caused friction among groups and led to policies of segregation being pursued by Whites in respect of other race grou~. Traditionally White South Africans have re sisted any trend towards the merging with Black and Coloured gr'oups and Afrikaners have also resisted fusion with the English speaking section, fust by the development of racial and ethnic at titudes which prevented meaningful social contact acrcgg colour and culture lines and later by legislation. The whole fabric of separate development, whether at territorial, occupational, social or cultural level, is aiIned at resisting the countervailing trends which have resulted from industrialisation. There is even legislation calculated to perpetuate cultural diHerences between Afrikaans- and English-speakers, e.g. the Provincial Kduca tional Ordinance requiring "home language" instruction. Demographically the diHerent groups have differing rates of increase, and there are signs that the rates will become increasingly unfavourable to the Whites in the years ahead. (Leistner m Rhoodie (ed.), I972: 2g9). THE KLITES LOCAT ION X% THE %W~

GR O UP

For the purposes of this study, elites were defIned as those persons at the top of the Volte occupational hierarchy. Since South Africa is a rapidly industrialising country, many elite members have risen to prominence from relatively humble beginnings, C onsequently, they can in large part be viewed as on~ d i n g citizens rather than as stemming from a distinct privilege class. A comparison of the elites and the larger White population in terms of such characteristics as geographic area, age, education, income and home language beaIs this out. Although it is true that the two groups diHer in a number of respects, these diHer ezIces are primarily the result of divergent career patterns rather than background or social factors. The geographic distribution of the elites does not diHer much from that of the larger White population. There is a tendency for the Transvaal to be ever-represented among the respondents (58 per cent of the interviewees come from this area as opposed to Q per cent of the population). This is, however, due to the fact that the civil sernce, a large portion of the elites, is concentrated in Pretoria and the headquarters of South Africa's major industrial concerns, as well as most national organisations, are located itt

A PRoFILE QF THE wHITE soUTH AFRIcAN ELITEs

g5

Johannesburg. The other provincial proportions are more equally represented in b oth el ites and p o pulation, as th e f o Howing figures indicate: 2 4 per cent o f t h e elites live in t h e C a pe, as

compared to gR,5 per cent of the larger White group; for Natal the figures for the two groups are i o and I I per cent respectively;

and the Orange Free State claims p per cent of the interview sample, as well as g per cent of the White population. The age distribution of the elites differs more sharply than that of the larger White population than does the geographical distribution. While 5y per cent of the White population are under 5o, only So per cent of the elites interviewed fall into this category. Elites, in fact, tend to cluster in the 5o-59 age bracket (47 per cent coinpared to Ri per cent of the population). This age differential is to be expected, however, in view of the fact that elite positions are achieved normally only after a long period of learning and experience in any given field. In his study of federal elites in the U.SA., W. I.loyd Warner ( i965: i og) has demonstrated the importance of education in the achievement of elite status by contrasting their educational level

with that of the general population. Data from our study indicate that the pattern in South Africa is similar. Only i8 per cent of the total W hite population have quahfications beyond high school, but 8o per cent of the eHtes have such qualifications, Similar to education, the average income of the elites is far above that of the general population. In i 9 68, I I p er cent of the elites earned more than RRo ooo p.a., 26 per cent earned between

Rio ooo and Rco ooo p.a., 49 per cent earned between R6 ooo and Rio ooo p,a., and only ig.5 per cent earned below R6 ooo p.a, In contrast only g per cent of the economically active White population in I96o earned salaries above R6 ooo (S.A. Statistic Pear Book, I966: H-28 ) while 87 per cent of the eHtes fell into this category. Although these figures are based on different years and are not strictly comparable, the gap is striking. In sum, elite members tend to be older, more highly educated, and better paid than their feHow countrymen. These differences between elite and population are, however, to be expected in terms of the two groups' differing occupational status and are not reaHy signs of sharp social or background cleavages. This conclusion is supported by the home language distribution of the two groups, According to the i96o census, the White population in South Africa is 58 per cent A frikaans-speaking, 3y per cent English-speaking, and i,6 per cent both Afrikaans- and Englishspeaking (see table R). The language distribution of the elite sample is not quite equivalent in that 4g per cent of the respon-

Q6

A PRO F I LE OF THE WM T E SO UTH AFRICAN ELITES

dents are Afrikaans-speaking, 45,5 per cent English-speaking and Io per cent speaking both languages. TABLE

2

Elites and White PaPulatunt bp Home Language

Home Language A&ikaans .. English Both Afrikaans and English Other Total

N

Elites p/p

%white Population, 1960 p 0 N /p

3 99 4 21 95 10

43, 2 45, 5 10, 3 1,0

1 791 1 151 45 102

58,1 37,0 1,6 3,3

9 25

1 0 0 ,0

3 089

100 , 0

The elite, as compared to the larger population, has a disproportionately large bilingual group which probably emerged in response to occupational demands and high educational achievement. In addition, the Afrikaans group tends to be under-represented in the elite sample, while the English group is over-represented. This is not surprising, especially in view of the traditional dominating position of the English-speakers and of the process of anglicisation that Afrikaans elites have undergone (described later in this chapter). In fact, the ethnic groups into which elites were born represent the language distribution of the population more closely, namely 5I per cent Afrikaans and g5 per cent English. These language diHerences between elite and population are relatively minor, however, especially when compared to the two grou~' educational or income divergences. This implies that elite and population differ not so much in terms of social background as in terms of those characteristics associated with t h eir respective occupational statuses. Consequently, the cleavage between the tIvo groups is not as pronounced as is the case in many other countries where elites are recruited from a distinct social class. M AJOR CHAR A C T ERISTICS OF THE ELITE LANGUAGE GRO U P S

Due to the polanmtion that has characterised the relationship bettveen the Afrikaans- and English-speakers throughout South African history, it is not mrprising that home language provides the key to most of the internal differences within the elite group be they attitudinal or otherwise. To highlight the clearcut associa-

tion between home language and other social characteristics, table 3 presents the pattern of social attnbutes linked to each of

A PROFILE OF THE WHITE SOUTH AFM C A N ELlTES

27

the three major language groups. It shows that ethnic background, religious preference, and political preference are strongly differentiated in terms of home language, while occupational sector, income and travel abroad show a more moderate degree of association. Age, education and contact with o ther race groups are not very closely associated with the language groups, but they are such basic attributes that they deserve brief inspection. Two other key social characteristics, sex and marital status, are not included in the table because they are so uniform. Almost all of the respondents (97 per cent) are male and 9 out of to are mat Tied (of the unmarried, z per cent are divorced. E per cent are widowed and g per cent are single). The ethnic background distribution for the elite sample as a whole reveals that g t per cent of the respondents were born in Afrikaans homes, 3g per cent in English homes and 8 per cent in homes where both languages were spoken. In addition, g per

cent of the interviewees come from Jewish homes (t t of the g8 respondents come from homes that were not "purely" Jewish) and 2 per cent come from homes where languages other than Afrikaans or English were spoken.+ Table g indicates that ethnicbackground and home language go hand in hand, as one would expect. Of the Afrikaans-speakers almost all are of Afrikaans parentage, whereas few of the English-speakers are. Almost three-quarters of this second group have parents who are of English descent. In fact, it is probably most useful to think of ethnicbackground and home language as measurements of the same variable at different points in tune, Ethnic background refers to language group a respondent's parents belonged to, while home language specifies the interviewee's present "ethnicity" or the language spoken in his own home. The ethnic background cieavages are closely paralleled by religious difkrences. The elite sample is composed of primarily two religious groups. Fifty per cent of the respondents belong to theDutch Reformed Church and another 86 per cent belong to other Protestant churches (primarily Anglican' ). In addition, a small Catholic group accounts for g per cent of the sample, 5,g per cent of the respondents are Jewish, and 6 per cent of the respondents stated they had no religious preference. These last three groups are of little consequence numerically and tend to be primarily English-speaking. The major language cleavage +For the purposes of analysis the small " Jewish" ethnic group and the "Other Language" group wire combined with the respondents stemming from biTingual homes m form a "Mixed" group (Le., ethnic backgrounds which are neither purely English nor purely Afrikaans).

28

A PR O F I LE OF THE WHIYE SOUTH AFRICAN ELrrES

is therefore in terms of the Protestant groups. Eighty-three IIer cent of the Afrikaans-speakers belong to the Dutch Reformed Yna t. u 3

Selected BackgroundCharacteristics by Home Language«

Background Variable

Home Language

Degree of Afrikaans Afr.-Eng. Assoc. with English (Per cent) (Per cent) (Per cent) Home Lang.oo

Ernie Background Per cent of Afrikaans parentage

997

57,1

Religious Pref. Per cent b elonging to D.R.C...

83,4

12,9

Polit& al Prej'. Per cent favouring National Party

51,6

5,7

58,4

49,0

10,3

235

p ( 0,001

V = 0,872

8,7

Sector

Per cent employed in the Pohty Per cent employed in the Economy

Tau = 0,821

Tau = 0,608

p ( 0,001

28,6

Income

Per cent earning below

RIO 000..

7 recce/ Abroad Per cent never travelling..

74,0

29,4

20,4

49,8

Tau = 0,240

10,1

Tau = 0,254

Contact toitft Other Race

p ( 0,001 p ( 0,001

CrouPs Per cent having frequent contact ..

Tau = 0,153

23,4

p ( 0,001

Education Per cent having more than a B.A.

42,8

Tau = 0,156

Age Per cent under 60

76,5

Tau = 0,059

Nutnber of cases on which percentages are based ..

(377)

(98)

p ( 0,001 p ( 0,01

(415)

«The tables presented in this chapter emerged as part of a more detailed analysis of the ehtes' images of the future. Consequently those respondents who did not answer the questions on the future were eliminated from consideration, thereby reducing the sample size from 924 to 888. The sample size was further reduced in specigc tables by the elimination of respondents who did not reply to the specific itetn being analysed. oo The statistics used are Cramer's V and Kendatt's Tau.

A PROFII.E OF THE WHITE SOI;TH AFRICAN ELITKS

g9

Church. In contrast, almost 96 per cent of the English-speakets b elong to o ther P rotestant groups, the largest of w h i c h i s t h e

Anglican Church. Political differences are less pronounced than the other ethnicaBy-associated cleavages just described, but they are still quite strong. Taken as a whole, ahnost half of the resp ondents favour t h e N a t i onal P a r ty , E I p e r c e n t f a v ou r t h e m iddle-of-the-road U n i ted P a r ty , I 6 p e r c e n t a d h er e t o t h e

liberal Progressive Party and Ig per cent do not prefer any of the major established parties. The major language differentiat ion in t erms of political preference is with respect to t he National PaIty; 93 per cent of the Afrikaanssupport it, while very few of the English do (only 8 per cent ). The latter are most likely to support the United Party (g8 per cent do}, although 33 per cent favour the Progressive Party. Of the respondents adhering to these two opposttion parties, over 8O per

speake rs

c ent are E n g lish-speakers. Finally, a bout tt o p e r c e n t o f t h e

English-speaking respondents stated they had no political preference, as opposed to 3 per cent of the Afrikanets. In sum, the ethnic background, religious preference and political preference of elite members are strongly linked to their present ethnicity or home language. Th~ c l eavages within the elite sample are typical of those found in the larger White population. Although home language is not as strongly associated with occupational sector, income, or travel abroad, as it is with the variables just discussed, meaningful differences do emerge. The interview satnple as a whole has the following occupational distribution: Polity Economy .. Education and Research Mass Media National organisations Religion Professions

42,8 20,5 9,8 7,8

7g

6,5 5,4

Total

Number of cases on which per cent is based

..

(888) The two largest occupational sectors serve to differentiate the language groups most clearly. Although over half of both the and Enghsh groups are employed in the pohty the economy, g8 per cent of the former cluster in the political sector as opposed to E/ per cent of the latter. Economic pursuits, on the other hand, occupy only Io per cent of the Afrikaans-speakers, but a9 per cent of the English-speakers. The

30

A PROPlLE OF THE WEfIX SOUTH AFRlCAN ELITES

second group is also far more evenly dispersed among the various sectors because, while the Af rikaners are disproportionately in the political sphere only, there are twice as many English- as Afrikaans-speakers employed in the economy, mass media and professions. These occupational differences are reflected in the income distributions of the two major language group. Due to their concentration in the economic spheie, the English-speakers tend to have a higher income level than the Afrikaners. Of the latter ahnost three-quarters earn less than Rio ooo ($iy ooo,', whereas only half of the Englishspeakers earn less than Rro ooo. For the sample as a whole, 6a per cent earn less than Rio ooo and, consequently, though the hnguage groups differ in terms of income, this cleavage is not pronounced. The same holds true for degree of foreign travel. Regardless of home language, close to two-thirds of the elite respondents seldom travel abroad (i.e., once every few yels). Inaddition, 2o per cent have never been abroad and i z per cent travel frequently. It i s i n t h ese latter groups that language differences emerge. Table 3 indicates that 29 per cent of the Afrikaners have never left South Africa, as opposed to io per cent of the English-speakers. In contrast, i7 per cent of the latter travel abroad frequently, but only 5 per cent of the former do. In comparison to the Afrikaners, therefore, Englishspeakers tend to travel more frequently, enjoy a higher income level and are more likely to be engaged in economic or professional pursuits. The Afrikaneis. on the other hand, tend to work in the political sector, earn less than Rio ooo and seldom, ifever,travel abroad. In contrast to the variables discussed thus far, home language has only a minor link with age, educational level and contact with other racial groups. The slightest relationship is with age in that abnost half of the elite interviewees are in their fifties,

represe nted

while 3o per cent are younger and somewhat under a quarter

are over 6o. This basic piofile ts essentially the same for all three language groups, as table 3 indicates. The educational structure of the sample (see chapter 3) is somewhat more comply t han is the case for age. The broad educational profile seems to emerge most clearh, however, when two educational categories are employed: Bachelor's degree or lms ..... . Higher degree ..., . . . . . . . . . . .

6 o p er cent yo p e r cent

T here is relatively little difference between the Af~ - and English-speakeis in t e rms of e ducational level, though t h e

A PROFILE OF THE WHITE SOUTH AFRICAN ELITES

QK

former are somewhat more likely to hold higher degress than the latter.

This slight educational differential may partially stem fmm the fact that the Afrikaner has experienced more upward educational mobility than his English-speaking counterpmt. Mobility was measured by comparing the respondents' educational level with that of their fatheIs' in tertns of seven categories. For each educatiozml step above his father that a respondent went, a. score of + I wa s awarded, and for each step below, a score of — r. Any individual's score could, therefore, range from to + 7. Since most fathers had a relatively low educational level — well over go per centhad no more than a B.A. degreethere was almost no downward mobihty. (Only 6 per cent of the respondents exhibited such a pattern). Consequently, only the positive scores were of importance and these could be grouped as follow~: No mobihty (negativedi8erence or ddference of 0) Moderate mobihty (difference of + I or +2) Pronounced mobiTity (difference of +3 or snore)

I00,0 Number of cases on which per cent is based

(779)

The above distribution reveals that almost four-fifths of the respondents have experienced some positive mobility. Of those who have not, the majority (6o per cent) are English-sp~ Only gu per cent of the English have experienced pronounced educational mobility as opposed to over half (56 per cent to be exact) of the Afrikaans-speakers. This differential mobility, which is reflected in the educational differences observed earlier, is probably a consequence of the greater value placed on formal education by the Afrikaans culture. Although the language groups differ relatively little in terms of contact with A f ricans, Asians and Coloureds, the divergences are worthy of brief mention. Respondents were questioned about contacts with other racial groups in some detail and their answers provided the basis for a G u ttman scale measuring the degree of contact elite members tended to experience regularly. Such interaction falls into three broad groups: occasional occupational contact (slight ); frequent occupational

32

A PROFILE OF THE WHITE SOUI'H AFRICAN ELITES

contact (moderate); and interaction in other contexts, in addition to frequent occupational contact (frequent). The elites are distributed within these groups in the following manner: Per ant

19,3 49,7 31,0

Slight Contact Moderate Contact Frequent Contact Total

100,0

Number of cases on which per cent is based

(886)

This general pattern remains more or less intact regardless of home language. Within each of the three language groups about go per cent of the respondents have moderate contact; that is, they engage in frequent interaction with other racial groups in their occupational settings. There is, however, a tendency for the English-speakers to be more bkely to have inter-racial contact outside of the occupational sphere. This is illustrated by the fact that 37 per cent have frequent contact as opposed to s'3 per cent of the Afrikaans-speakers. Still, by and large, the language different in terms of contact with other racial groups, are as limited as those for age and education. SOME CAUSAL PATTERNS+

Those variables which are most closely related to h o m e language in table 3 (see right-hand column) are also those which have the greatest impact on various elite attitudes, as illustrated

in more detail in chapter g. That is, the four background variables that consistently differentiate the elites' attitudes on various issues are ethnic background, home language, religious preference, and political preference. In c o ntrast, variables that

are less closely related to home language have a weaker, and less consistent, i m p act o n

t h e a t t i t ud e v a r iables investigated.

Variables measuring such things as occupational sector or degree of foreign travel, for example, tend to interact with home language to inHuence only a few attitudes. The last three variables listed in table 3 (i.e. contact with other race groups, education and age), which are most weakly related to language cleavages in the sample, seldom serve to differentiate elite attitudes. Since the first variables listed in table 3 are so much more «For a more detailed analysis of the causal patterns discussed in this sec-

tion, see Huber (1973: 63-78).

A PRoFILE oF THE wHrrE soUTH AFRlcAN ELITEs

33

closely related than the others, it is useful to consider briefly the causal relationships that prevail between them.

The causalhnks between ethnic background, home language, and political preference appear to be well-defmed by temporal considerations. For ethnic background (i.e., language group of parents) is given at birth, a home language is acquired somewhat later (although it sometimes changes), and pohtical preference develops last. This iinplies that the three variables are related in such a way that the first variable shapes the second which, in turn. moulds the third. That is, ethnic background influences political preference through its i m pact, on home language. As a result the last becomes an intervening variable and controlling for it should cause the relationship between ethnic background and pohtical preference to fade away. Analysis proved that this is just what happens. The original relat ionship between the tw o v a r i ables is quite strong i n t h a t 8 6

per cent of the born Afrikaners prefer the National Party, while only y per cent of the English-born do. Of the latter, y2 per cent prefer the Umted Party and 34 per cent favour the more hberal Progressive Party. In addition. the respondents of mixed parentage have a pattern similar to the English one. When home language is controlled for, the relationship between ethnic background and political preference just outlined, is sharply reduced. It tends to go to zero within the two major language groups, which means that home language intervenes in the re-

lationship between ethnic background and political preference. In addition to being influenced by ethnic background and home language, pohtical preference is also afFected by religious preference, a commitment developed early in every South African's life. The relationship between the two variables is pronounced, as 9o per cent of the respondents belonging to the Dutch Reformed Church support the National Party, whereas few members of other religious groups do. The latter tend to favour the U n ited and P r ogressive Parti' .

I t i s q u i t e possible,

however, that this relationship is largely spurious, for the variation in both variables might be due to the impact of home language. Analysis revealed this was not the case, however, for although controlling home language reduces the relationship between political and religious preference, a considerable portion of it remains intact. This implies that religious preference has some direct impact on political preference, even though both variables are in part defined by home language. In addition, political preference and religious preference, as well as

A PROFILE OF THE W'HITE SOUTH AFRICAN ELITES

home language,are inHuenced by ethnic background. Diagrammatically, the overall causal sequence that emerges is as follows:

Home Language Ethnic Background

Pohtical Preference

Preference THE ANGLICISATION OF AFRIKANER ELITES+

All of the variables discussed in the preceding section are closely associated, but the strongest direct link is between ethnic background and home language. In fact, as has been noted earlier, it is probably most useful to think of the two as measurt. ments of the same variable at different points in time. Table 4 indicates that although ethnic background and home language are identical for most respondents, there is some change in ethnic identity through time, especially among respondents of mixed parentage.As measured by home language, about I9 per TA s x. a 4

Honte LangnagcJ' o EhfcsbpEthnic Backgcound (Pcrccntagcs)

Home Language 82,6 5,1 12,3 Total

723 277

0,3 96,7 3,0

100,0

Number of Cases Zend-Order Relationship — Kendall's Tau = Controlling for Age: Under 50 .. Tau = 50-59 .. .. Tau = 6 0andover . . T a u =

0,821 (p < 0,001) 0,833 (p < 0,001) 0,816 (p < 0,001) 0,814 (p < 0,001)

oror a xnore detailed discussion of change in ethnic identity among the elites, see Huber (l973: 78-85).

A PROFILE OF THE WM T E

SOL~ A F R I C AN EXZrZS

cent of the respondents prexndy identify with an ethtuc grQup other than the one into which they were born. Almost all of these respondents have moved closer to the Eng away from being lish haguage in that they have p ~m A f r ikaners (or of mixed parentage ) to speakmg both Afri kaans and English at home, or just English. It appears th «re, that the Afnkaners are steadHy losing ™ i por t i o n of their Rock to the Enghsh. Respondents who are English born ahnost never become biTingual, let alone exclusively Afrikaans speakmg, within the con6nes of their homes. The fear of such anglicisation has haunted the Afrikaans community for decades. The superior position of the Knglishspeakers in the spheres of education and economy during the early part of t his century caused some Afrikaners to prefer English as a medium of instruction even in their own schools and its use in most walks of life. In his classic dissertation on the rise of the Afrikaner in the city, S. Pauw describes situation as "spiritual slavery" of the Afrikaners. (Igy6: Iy@) If it were true that the dominant position of the English corn munity in the past was largely responsible for the proc(Es of anglicisation, but that this pattern is changing, one would expect the relationship between ethnic group and home language to increase mth decreasing age (e.g., a smaller proportion of respondents would have changed theirhome language in the younger age groups where the influence of the English culture Qu]d have been less preponderant). But even when one controls for age, the relationship be@veen ethnic group and home language remains mrtually the same (see the statistics at the bottoIn of table @). It appears, then. that anglicisatiQn is not a thing of the past, but is still continuing. People of Afrikaans descent who adopt English, or English in addition to Afrikaans as their home language, do not mam the same traditional links with the Afrikaans community. There seems to be a trend towards favouring functional reLL pons with the English language group even at the expet~e Qf traditional links. Munger refers to this Afrikaans group which speaks English at home as the "Anglicised Afrikaners". He that they are "socially, economically, and poiiticaiiy d ivQJ'ced from Afrikaner nationali~ " . "They claim tQ South Africans and not Afrikaner or English, but the group has httle cohesiveness.. . T h ey are South African in the absence of being anything else, but their position tends to be amo(phous. fheir only unity is in reaction to criticism". (Munger, Ig67: 22). Consequently, it is of intetest to consider brieRy thee fac

rogram ed

ee

36

w eRo r ~

o F m z w H r r x sotrrH s t u m @Kt.tYas

tors that might have caused Afrikaner elites to change their home language to English, thereby exchanging the security of the Afrikaans heritage for the ambiguity of rootlessness. Of the background variables on which information was available for the elite respondents, three might have a causal impact on the relationship between an interviewee s ethnic background and home language: childhood locale, high school medium of instruction and Ml lvefsttyexpeflence. Analyses of these variables indicate that Af r i kaans-born res-

pondents, and those of mixed parentage, who have changed their home language are more likely to have been exposed to English environments during their formative yeats than are their Afrikaans-speaking contemporaries. Growing up in English or foreign communities, amending English schools, or going to English universities, as well as foreign ones, seems to be conducive to change in home language among Afrikaners. Among the English, in contrast, exposure to Afrikaans influences during youth seems to be much less important and does not lead to the abandonment of the mother tongue. It should be noted, however, that Afrikaners are exposed to a more diverse and less homogeneous enw~nment. That is, A f rikaners are quite likely to come into close contact with English culture during the formative years, while the English are not usually exposed to Afrikaans ways. It seems, therefore, that although the Afrikaner may be more affected than the Englishman by cultural influences emanating from other ethnic groups, the former is also more likely to b e exposed to such i n f l u ent than the latter. Change in home language, then, is not only the result of susceptibihty, but also a consequence of exposure. So far we have dealt only with factors that might precipitate a change in home language. It now remains to consider whether such ethnic change is purely lingual or more far-reaching as Munger claims. The factors of greatest relevance to this second concern are r e ligious preference and political p r eference. Analysis of these variables shows that re-pondents who have changed their ethnic identity are located between the two major language groups in their religious orientation — somewhat over

a third belongs to the D.R.C., while 6g per cent belong to other church groups. On the whole, this pattern seems to be closer to the English one than to the Afrikaans one. The same is true for pohtical preference. Only 3 t per cent of the group changing its language patterns belong to the 'National Party, while close to haM of the "language change" respondents belong to the United or Progressive parti' . T h ese political and religious patterns

A PROFILE OF THE WHITE SOUTH AFRICAN E~

S

would seem to indicate that, change in ethnic identity involves Inore than simply learning a new language. Afrikaners who become biTingual at home or English-speaking do not remain Afrikaans in other respects. Rather, they exhibit religious and political preferences typical of the English group. The much weaker trends exhibited by the language groups in terms of occupational sector, income level and travel abroad tend to re-con6xm this conclusion. As Afrikaners abandon a unilingual orientation, they tend to be more likely to be emsectors, more likely ployed in the economic or p to earn over R I o ooo and more likely to t ravel frequently. These traits are all more typical of the English-speaker than of the Afrikaans-speaker. This is not to say that language change causes the Afrikaner to change his patterns in terms of t he variables just mentioned. Rather, both t >yes of change probably

rofessia oln

p roceed simultaneously and r e-enforce each other. I n

f a ct, it

may well be advantageous to view the relationship between ethnic background and present home language as an indicator of the extent to which an individual has changed a whole complex of ethnically associated attitudes and attributes (some of which have been mentioned above) through time. This implies that there is more to the Afrikaner's speaking English at home than mere language proficiency. It appears, instead, that the larger English culture has an attraction for a segment of the Afrikaner group, while the Afrikaans culture has little to recommend it to the English-speaker. This may be because the South African English sub-culture has grown up largely in urban areas and has thrived on the early stages of local industrialisation. Consequently, it is better suited to the d emands of th e m o d ern W estern li fe-style than t o t h e m o r e

rural and provincial Afrikaans culture. Furthermore, the process of "anglicisation" may well increase as South Africa becomes more industrialis&, thereby having a substantial impact on the future of the Af rikaans culture.

Viewed in this light, the preceding discussion has a direct bearing on the dispute p raging within the National Party about how best to preserve Afrikaans traditions. It app ears that the conservatives' claim that interaction with t h e

resently

E nglish leads to m o d ification of t h e A f r i k aans culture i s a n

accurate one. Those rngondents in th e sample who h ave changed their language and ethnic identity have indeed given up the "Afrikaans way", especially in terms of religious and political preferences. It would also appear, however, that these respondents have done so because they 6nd the old traditions

38

A PRQFILE OF THE wHITE soUTH AFRIOAH ELITEs

of little value in the urban industrial environment. Unless, therefore, the Afrikaans culture comes to terms with the demands of the modern world, it may be weakened by the process of attrition. This may well occur if the Verkramptes' isolationist stance prevails and the Afrikaners withdraw to the safety of their rural "laager". If th e A f rikaans culture is to thrive, it will have to adapt itself to the demands of a modern industrialised society~ son1ethlng that can only occur through interactIon wtth

the more urbanised English and the outside world. It is this course of action that the Verligtes are recommending. Paradoxically, it seems that the preservation of the Afriltaans culture lies not in keeping it pure through isolationism, but in broadening it through exposure to new influences of the modern world. In the process many old traditions will be lost, but at the same time a stronger cultural fabric may emerge.

SOCIALISATION OF THE ELlTES

C hapter v highlighted some differences between the t w o language groups. It is shown in several chapters that follow that English- and Afrikaans-speaking elites di8er markedly with respect to the attitudes they hold towards social and political issues. The purpose of this chapter is to describe the socialising experiences of the elites in terms of the broad language division in the hope that it might shed light on the attitudinal differences referred to. The background characteristics dealt with are place of birth, childhood locale, educational institutions attended and religion. The formation of attitudes stretches over a lifetime and attitudes are the product of varied sociahsing influences. Other elite studies, e.g. Matthews ( i95g), I aswell and o r n e r ( ig65), Edinger and Searing (i967), have organised background variables to form what might be described as a "social profile" (defined by Janowitx (tgg6) as "the uniformities in patterns of social backgrounds" ) and have drawn attention to the relevance of such profiles for elite attitudes and behaviour. The attitude patterns of ehtes showed marked contrasts when analysed according to the language group of the people concnncd. What follows is an attempt to describe the contrasting background differences of the language groups and to suggest an interpretation of attitudinal differences in the hght of these factors with reference to supporting evidence from other studim. Hess and Torney (i 96): 9g) etnploy a three-fold classification of sociahsing mfluences in their study of political socialisation. The first are those influences which are exerted by weB-defmed structures and organisations and can be specificaHy identified. In this chapter, education falls into this category. In addition, data on religious backgrounds of the respondents' families of origin are presented and put into the same category because it is a pronunent influence in a well-defined institution, the family. The second category of socialising in8uences are those which take place within a wider social setting, where they cannot be

$0

SoctAusATION AND EDUCATION

defined precisely, but are numerous and subtle and woven mto the texture of everyday life. Country of birth and childhood locale are placed in this category, Cluldhood locale refers to the early environment of respondents and is described both in terms of its rural or urban character and also according to its predotmnant c u l tural c h a racter, i ,e. whether Afrikaans, English bilingual (i.e. both major influences). Social class, an important influence in the larger setting is not described. This is because the great majority of respondents in all language gro ups ha fathers whose occupations feil into the middle class white-collar category (see chapter g). As a factor of socialisation, therefore social class could be said to be "constant" in this study. The t hird category concerns personality, and is not r elevant to t h i s study.

Place of birth Eighty-six per cent of elites were born in South Africa, 9 per cent in the United K i ngdom, Ireland, Scotland, and other areas

of the British Commonwealth (excluding Africa), t per cent in Africa outside the Union, and g per cent in Europe. Other studies, e.g. Rogers and Frantz ( t96g) and Pettigrew (t958) have found that people born on the African continent were more conservative than those born elsewhere. Abl est all the Afrikaans-speakers and bilingual elites were born in South Africa, whereas the proportion is lower among the Enghshspeakers, a noticeable minority (gg per cent) ol whom were born abroad, Father's place of birth is important, too, and the number of English-speakers with foreign-born fathers is than is the case among the other two groups, Among Enghshspeakers, having a foreign-born father can be associated with the holding of liberal attitudes (Lever and Wagner, >965). Chi Mhood locale The predominance of rural backgrounds among Afrikaansspeakers indicates a strong predisposition to attitudinal conservatism. (In South A f r ica the ' platteland" has traditionally h ad t his reputation). I n a d d i t io n t o th i s , most A f r i kaners were

reared in an Afrikaans en>monment, implying that for the m« part their peer group were Afrikaans-speaking like themselves

and that they w ere not exposed to a w i d e r ange of contracts

Both the Knghsh-speaking and bilingual groups gave evidence of a greater variety of early peer group contacts and far less evidence of a rural influence (see table 5 ).

SOG1AI.ISATION AND EBUCAYION 'rA S I . E

5

Chifdhoodloeofe by Language Group oJ Elires (Pereentuge)

L ANGUAGE GR O U P

CHILDHOO D LO C A L E

A&ikaans K = 399 A&ikaans Rural Afrikaans Urban English Rural .. English Urban Afrikaans-English Rural and Urban Outside South A&ica Total

Knghsh 'A = 422

Bilingual K = 95

47 I850 lO I

22 l6 IOO

IOO

High school A school offers lwo categortes of soctaiisation children. The first is that which is deliberately aimed at and created by the teachers and policy-Inakers of the school. In this situation children are consciously and deliberately taught things f or a speclflc reason. The second Is the result of c o ntact w i t h

people in the school. The composition of the staff and pupil body, the latter comprising an important part of a child *s peer group, will have an important effect on the acquisition of attitudes. Contact with social groups other than his own in the school situation is an important socialising influence in a child's life In general, schools which have a homogeneous student body are likdy to reinforce group norms as the students are never exPosed to contact I~~th members of other grouPs at school. This applies to the pupil body as well as to the teaching staff. The English-speakers reveal a school background which reHects the fact that only a very smail minority had any contact with Afrikaans-speaking children at school tsee table 6 ). A mere I I per cent appear to have had contact of this nature. The

rat went either to South African Enghsh schools or to schools overseas. Most of these latter were British. The most important

educational institution in this group's education is thus the English South African school. These schools have no strong and positive educational philosophy equivalent to that existing m Afrikaans schoob. GeneraHy, the schools could be said to r e Hect a desire to foster a broad South Africanism. This implies

SOCDlLISAITON AND EDUCATION

the inclusion of all the Whites in one South African nation. Exactly how far they would be prepared to go and, more importantly,would have been prepared to go go years ago when our respondents were in attendance, is a moot point. It can be said, however, that they have not set out consciously to foster

the nerce pride in group identity that their Afrikaans-medium counterparts have. English-speaking South Africa. has never had a strong sense of its own identity as a group. Nevertheless the schools offered little opportunity for meaningful social intercourse with members of the other White language group. The number of Afrikaans children who have attended English-medium schools is very small. As far as Africans, Coloured people and Indians are concerned, the schools do not really do anything positive about their pupils' racial attitudes one way or another. The only contact they would have in the school situation would be with cleaners and gardeners at the master-servant level. I A S X. S 6

High Schon/ AttendedAcccrding «sLanguage Gmup (Pcrccntugcs) HIGH SCHOOL

L ANGU AG E GROU P Engliish K = 422

A&i%sans English Both (Bilingual) Foreign Total

51 13 34 2

Bilingual N = 95

4 73 7 16 100

The schools which teach through the medium of Afrikaans have a much betterdeHned and determined purpose. They have been assigned a very important place in the cultural scheme of Afrikaner life. Christian-National education involves a total approach to education, in which all subjects are taught, all activities co-ord inated with the same object in v i ew. T his i s t o p r o duce the Christian Afrikaner, loyal to his people and an upholder of Calvinist principles. It is not possible to say with any accuracy to what extent this was applied in schools attended.

by our respondents, but it ce~ ly c a n be said that in many the inHuence must have been a very stttsng one. It was, how-

SOCIALISATION A~

E D U CATION

ever, not oScial policy anywhere, although the rural areas must have been particularly affected, especiaHy in view of the role of the Church and the "prahkant" (minister) in local affairs. The religious influence would be particularly strong on the school committees and school boards. Each school has a school committee which selects and appoints teachers and has limited power in the school it+:If. The school board ratiftes appointments. Pupils at A f rikaans-medium schools would have been even more isolated in their social contacts than those at English-medium schools. We now come to the bilingual or parallel-medium school, that which teaches through both languages. It would appear that this would be the place where a common White South Africanism would be the prevailing spirit, the presence of both language g roups making for c loser contact among children. I n t h i s arrangement there are parallel classes for the different language groups, which means that chfldIen are taught in their home language, but are likely to be exposed to both languages during extra-mural activities, etc. Another system known as dual-medium, by w h ich teachers switch from on e l anguage to another

during the course of a lesson has, for practical reasIns, never been very successful. T hese paraHel-medium schools were very important in t h e education of our Afrikaans-speaking respondents, gg per cent of w hom attended them. Furthermore, over one-third of t h e bilingual group also received their secondary education in these schools. What are the implications for socialisationP It has already been stated that Afrikaner socialising influences provide a unity of theme which is not present to the saIne extent in the English case. Clearly schools at which both groups are well represented would be places where attitudes would be formed amidst contrary influences. The flrst factor that has to be taken into account is the demo-

graphic one. Until fairly recently, the distribution of the White population was such that the urban areas were predominantly English and the rural predominantly Afrikaans. The schooling pattern that developed was that the schools in the urban areas were predominantly English and those in the rural areas Afrikaans. Thus by and large the schools were single-medium. This was p true in the cities because there would always be a number of high schools, mostly English, a few Afrikaans.

arti cularly

A n urban child w o ul d

t h e refore attend th e school of hi s o w n

language. It was in the country areas that the parallel-medium

SOCIALISA'rION A'VD EDUCATlON

schools were to be found. This situation arose because of the need to provide instruction in the mother tongue for the smaH minority groups who lived in the rural areas. In most cases there would be English children who would claim the right to be educated in Enghsh in some small country t o wn, thus turn-

ing the school into a parallel-medium one. In effect, however, the school would remain overwhelmingly Afrikaans in every respect. Thus for present purposes the parallel-medium schools are regarded as being in the same category as the Afrikaans singlemedium schools. The teaching staH wo »d be very much the same and so would the genera1 aims and spirit of the schoolsAs far as contact with the Enghsh group was concerned, it w ould have been restricted to a

t i n y m i n o r it y o f p u p ils w h o

would have had a very weak presence. In conclusion, the South African school system has never been conducive to blurring signi6cantly the lines of demarcation b etween gmups. M e a ningful

c ontact w a s d e n ied t o o u r t e -

spondents, when they were at school, with members of the o« r language group. 'nevertheless the attitude of the English-speakers in this study are more open and favourable to the idea of c lose contact wit h

A f r i k aners than v i c e v ersa, an d a f a c t o r

which conceivably contributed to this was the less pow«» emphasis placed on group identity in the English schools. Uni venif y Table 7 presents the data concerning the university attended by respondents in the different language groups during their undergraduate years. We have not included the institutions atten-

ded for post-graduate study at this stage because of the smaller numbers. It mn also be argued that the undergraduate years are more formative. From a n a t t i t u dinal p o in t o f v i e w , a t t itudes are less rigid in t h ese years than later and c onsequently more

open to change and impression. Thus the type of teaching that is received is of great i m p o rtance. This spirit t h a t m o t ivates the university. the role it sees for itself in the aHaus of mankind and its own social group in p ar ticular, are both extremely important in determining the academic ethos. In a d d i tion, contacts with

fellow students are usually at a much deeper and more sigruftcant level at this stage than later.

The pattern of attendance according to language group follows that of school attendance. The great majority of A f r ikaanss peakers attended A f r i k aans-language univetsities. A m er e t o

per cent were exposed to the influences of an English-language

SOCIALISATION AND EDUCATION T ABL E 7

Uralergrakcute Unitlersitp Attended by Language Group fa Elites (Percentages)

Language Group

A&ikaans English University of South A&ica

Foreign None Total

55 IO 9 l 25

100

utuverstty. The Umvetstty of South Africa aho claimed a small

minority, but it must be borne in mind that it has never been a residential university. It has always conducted its coutses and examinations by correspondence. In addition, in the early days of South African higher education it was the only degree-granting institution and acting as the examining body for all the university colleges it had to accommodate a wide variety of viewpoints, thus maintaining a considerable flexibility of outlook. It. is therdore extremely dificult to classify in terms of the EnglishAfrikaans dichotomy we are using. The Enghsh-speaking group reveaL~ a similar tendency to attend universities of their own language medium. Rather less went to Afrikaans universities than was the case with Afrikaners attending English institutions. Also, a not inconsidtxable percentage attended universities abroad. The bilingual group showed an expected distribution over both English- and Af~ -lan g uage institutions. Thus the university socialisation pattern does not reveal any deviation from the pattern so far. The language groups are again separated by their universities and proceed on their different way~.

It is broadly accurate to say that the English-language universities have reflected the ideal of teaching and research which is directed towards cosmopolitan concerns. There is little con-

sciousness of group and no firm conception of a specific role in the political sense. The Afrtkaans-language utuversities — SteHenbosch, Pretoria, Potchefstroom and Orange Free State — pmwnt a different picture as do th e more recently established ones, the

SOCKALISATION AKD EDUCATION

Rand Afrikaans University founded in I g67 and the University of Port Elizabeth founded in tg6g. (Though predominantly Afrikaans the latter teaches through both languages.) During the university careers of our respondents the latter two were not in existence and are thus omitted from the discusston.

While not unmindful of universal concerns, these universities are more bound to the local and particular. The Potchefstroom Univemty for Christian Higher Education is perhaps the most intensely "national" i n c h a racter and c a used Sheila Patterson to comment that student life was, "concerned with a particular

set of ideas and activities proper to the Christian National way of life". (Ig5p: u34). The other three Afrikaans-language univeIsities, even though less intense than Potchefstroom, would nevertheless reflect their Christian-National orientation in both their teaching and their research. Ahnost certainly student contact would be with members of one's own group only and, although this is also largely true in the case of the English-speaking students, there has been a token group of Africans, Coloureds and Indians on the English-language campuses to challenge traditional attitudes. In order to develop a more complete picture of the higher education pattens of the two groups, post-graduate and underg raduate institutions w ere coupled. E x a m i nation o f

data r e-

vealed that @I,5 per cent of the Afrika ~ p ea k e Is attended Afrikaans institutions only, both a t t h e u n dergraduate and post-graduate level. Thirty per cent of their English-speaking counterparts attended English South African institutions. Both language groups had a sizeable percentage who attended universities outside the c o untry — II

pe r c e n t o f t h e A f r i k a ans-

speakers and I6 per cent of the English-speakers. Thus again the same trend is observed. Enghsh-speakers were exposed to more heterogeneous educational influences than the A f r i kaans-

speakers. The picture for the bilingual group is one of attendance at both English and Afrikaans universities in substantial numbers.

ReIigion The religious influence a child is exposed to at home is clearly a factor which has to be taken into account when trying to understand the relationship between attitudes and the socialisation process. Table 8 reveals that the Af ~ group is almost completely (g4. per cent) the product of homes where religion of the head of the house was Dutch Reformed. It can

SOCIALISATION AIID EDUCATIO'S

be safely assumed, too, that there would be a few cases where a counter-influence was present by a mother having a different

denominational allegiance. This is of great importance. The Dutch Reformed Church has stood as a pillar of strength in the Afrikaans Nationalist movement. It has been a stronghold of Afrikaans language and culture and has lent support to the political aspect of the movement. In th e field of attitudes, it )ms

s tood uncompromisingly for t h e p r eservation of t h e A f r i kaans people through a rejection of intimate contact with the other White Groups, and aLm, Rhodes and Natal; the University of South Africa (UNISA), a bilingual correspondence institution, and the U n i versity of Port

Elizabeth catering for both major language groups, yet predominantly Afrikaans. Sixty-three per cent of respondents in the total survey were university graduates. As in the case of schools, most attended universities where their home language predominated. Thus of the ~g per cent of A f rikaans-speakers who attended university, g5 per cent were at Afrikaans institutions and of the 58 per cent. of the English-speakers who were graduates, gg per cent attended South African English institutions and approximately t 8 per cent British univetsities.

The University of Stellenbosch has educated more ehtes than any other South African university. 4'Vhen compared with the other Afrikaans universities, its position is particularly outstanding, v ith

t h e U n i v ersity o f P r e toria qu ite some distance

behind. The Afrikaans universities have played an extremely important part in equipping the Afrikaner population with the

SOCIALJSA'rLOA AND EDUCATIOK

skills necessary to assume their present dominant position. Thus they have been popular universities, this being particularly true of the University of Pretoria. Nevertheless the University of Stellenbosch, possibly on account of it being the longest established university, has acquired something of an elite reputation. It is di%cult, without f a t h er i nvestigation, to say anything about the relative importance of the Afrikaans universities in South African life, but Stellenbosch certainly has a high reputation as an institution wit h great cultural and political signifi-

cance. In addition, there are a high proportion of students in

residence, a fact which could be of signi6cance. In the case of the English-language institutions, the University of Cape Town has the predominant place, followed closely by Witwatersrand. The Enghsh-language universities are institutions existing primarily for the purpose of preparing students for careers in commerce, industry and the professions and have drawn their students from all strata of th e English-speaking South African population. Their values, traditionaHy approximating those of the Western academic world, have set them at variance with White South African political ideas (Ashley and Van der Merwe, t969) and they have consequently experienced something of a political and cultural isolation. Their character thus does not reflect what A . H . H a l s ey (t96t) has called the "English Idea*' of a university, i.e. regarding higher education as the preparation of future elites for their status by imparting a distinctive life-style, after the manner of Oxford and Cambridge. It is wo M n o t ing, however, that elites educated at British universities feature quite often in the education, religious and economic sectors. The large percentage of such graduates in the education sector could be ascribed to the fact that South African univnsities and, in p articular, the English-language institutions, have always looked overseas, and to Britain especi-

ally, for staff. In addition, a large number of academics attend foreign universities for post-graduate study. In the religious sector the Anglican church in particular has traditionally appointed most of its bishops from the ranks of English-born clergymen and there are i nstances of this i n o t her E nglish-language churches. The.e men have had their education at British universities, particularly Oxford and Cambridge. In the economic sector the explanation probably lies in the traditional contacts with the United Kingdom held by several extremely prominent industrial and commercial organisations. The mining industry and banking are particularly good examples. The contention that the University of Stellenbmch is an

6o

SOCIAI ISATION AND EDUCATION

elitist institution is strongly supported by other evidence from this study. were asked which contacts had been most helpful to them in the attainment of their present positions: (a', those made at school; (b) those made at university; (c) neither of these; (d) both of these. Sixty-two per cent of the Stellenbosch graduates said that their university contacts had been most helpful to them, as against @I per cent from Pretoria, gg per cent from the University of Pot-

Respo ndents

chefstroom and 29 per cent from th e U n i v ersity of the O r ange Free State. It would seem then, that while in a t tendance at the

University of Stellenbosch, maay of the persons induded in this study met and developed lasting contacts which aided them subsequently in the attainment of their positions. This is, of course, one of the characteristics of an elitist institution and th e same

pattern could be observed at Oxford and Cambridge and the Ivy League Universities. Schools did not play an elitist role in this respect, i,e. where lasting contacts were made, thus confirming earlier conclusions. Thus it would seem fair to d raw the conclu-

sion that the ruling Afrikaans elites in South Africa met and established contacts of great importance in the Afrikaans universities, particularly Stellenbosch, while Afrikaans schools played almost no part in this process. By contrast, the English-language universities are clearly not p t he same role as about a

laying

quarter of the graduates said that their u n iversity contacts had

been of importance to them. English schools were even less important. Respondents who were not university graduates did not affect the general pattern. Eleven per cent said that schools were

places where important contacts mere made. The White South African elites thus had a school background which showed little trace of the influence of self-consciously e litist institutions. In t h e c ase of th e A f r i k aners, the most im portant type of school was that situated in country towns, with

pupils drawn from a variety of social backgrounds. The urban state boys' schools were most important for English-speakers, with elitist private schools having a minor influence. The English-language universities were seen p r imarily a s t r a ining grounds for industry, commerce and the p r ofessions, being places of lunited importance for the establishment of significant personal relationships and not dedicated to imparting a specific Vife-style. The Afrikaans universities, Stellenbosch in particular, were places where important personal relationships were established. As to their political and cultural roles, and the nature of t h e s o cialisation p rocess students u ndergo, f u r ther investigation is necessary.

CHAPTER IV

THE RELATIVE IMPACT OF ETHI C A E D S1RUC1 URAL FACTORS OE AT1 ITUDES TOWARDS SEGREGA TIOX AED MSSEJVT Since the close of the Second World War, two issues have served to divide White South African society. One of t hese issues is the degree to which various racial groups ought to be separated by law and the other deals with the extent to which citizens ought to be allowed to disagree with government policy. These two questions lie at the root of a good deal of controversy in contemporary South Africa, and consequently members of the elites were asked to express their views on them in some detail. This chapter outlines the nature of their attitudes on these two divisive issues and tries to i solate the factors in-

Huential in moulding their views. It is shown that the most important factors are the ethnically-linked background variables. I n addition, occupational sector i s sh ow n t o

h ave a l i m i ted

impact on attitudes towards segregation. It emerges as a conditional factor which serves to exacerbate or mitigate the inHuence of the ethnic variables on attitudes in certain occupational groups. As South African society becomes increasingly industrialised and structurally complex in the future, occupational context will no doubt be more inHuential in attitudes. At present, however, the impact of sector is minimal in comparison to the ethnic variables. The pervasive inHuence of ethnicity on attitudes towards segregation is typical not only of the ehtes, but also of the larger White population as other researchers have shown.

moulding

ATTrrUDES TOW A R D S L E GAL SEGREGATION

The questionnaire item designed to tap elite members' ideas about the desirability of v a rious types of segregation was phrased as follows: "To what extent should sqnegation between Whites and other races be enforced in the following spheres 6t

ETHM C AND STRUCTURAL FACTORS

of lifeP" The twelve spheres to which the question applied ranged from toilet facilities and residential areas to labour unions, from universities and private homes to sports events. Respondents were asked whether segregation in e ach o f

t h e se

areas should enjoy rigid enforcement by law, limited legal enforcement or no legal enforcement at all. Of the original twelve spheres, eight lent theamlves to combination as a Guttxnan scale measuring respondents' general attitude towards legal amenable segregation. In order to make the original q to scaling, the following reformulation proved necessary: Should rigid legal segregation between AVhites and other races be enforced in the fouowing spheres of lifeP

uestions

(a) Toilet facilities

(b) Hotels (c) Beaches (d) Buses (e) Universities

(I) Churches (g) South African sports teams abroad (Ii) Visiting sports teams By modifying the question in this manner, the original three response categories were consolidated into two: "yes" and "no". The positive response applied to one of the original categories (i.e., segregation should be rigidly enforced by law ), wtule the negative answer included the other two (i.e. legal segregation should be enforced to a hmited extent, and there should be no legal enforcement of segregation). Scale scores were arrived at by combining the modified responses for each of the spheres listixl above. The resultant segregation index served to summarise a series of related attitudes pertaining to diverse spheres of life. (For a more detailed discussion of the scale's construction, see Huber, t9pg i Appendix III. ) Inspection of th e d istribution of scale scores shows that about two-thirds of the respondents favour some form of rigid segrq,ation (i.e., separation legally enforced without exception). The other third of the elites do not favour rigid segregation in any sphere,though only is per cent advocate no legal segregation at all. The rest favour limited segregation (i.e., while the races are legal]y separated, the law i~ applied leniently and exceptions can be made) in one or more spheres. The elites are least likely to favour segregated sports teams (only ig per cent do) which would seem to indicate that there is widespread sup-

ETHNlC AND STRUCTURAL FACTORS

port for the National Paxty's new policy of placing international sport beyond the jurisdiction of apartheid legislation. A considerably greater proportion of the elites stated separation of the races was desirable in other areas, however.

Since next to none of the original categories contained more than xo per cent of the respondents, they were consolidated into two larger groups for the purposes of analysis. The basis for differentiating these groups was the extent to which respondents favoured rigid segregation in the fixst four spheres (e.g., toilet facilities, hotels, beaches and buses ), or in the last four spheres (universities, churches and sports teams). The spheres within each of these two groups tend to cluster together in that if respondents favour rigid segregation in one or more spheres in the second group, they are very likely to favour it for all of the in the first group too. If , however, respondents favour rigid segregation for only some of the spheres in the first group, they do not tend to favour it for any of the spheres in the second group. When the various background variables were related to the segregation index, it became clear that it had much stronger differentiating power than many of the other dependent attitude variables examined. That is, respondents belonging to d i f ferent groups in terms of a given social characteristic, differ sharply in terms of their beliefs about the degree of rigid legal segregation that is desirable. Although the relationships of the background variables with the segrefmtion index are unusually strong, the general inter-relationships are much the same as those uncovered by analyses using more weakly related dependent variables. The pattern of relationships that emerged for attitudes towards legal segregation involves ethnic background, home language, religious preference, political preference and

spheres

occupational sector. These are the same variables that proved

relevant in analyses of other attitude variables, such as the images of the future discussed in chapter 9. The relationships between home language, political preference and the segregation index axe quite strong and dearcut as table g shows. About thrm-quarters of the Afrikaans-speakers favour moderate or pronounced legal segregation as opposed to in per cent of the English-speakexs. The vast majority (88 per cent) of this second group favours limited segregation, at most. The political cleavages are quite similar to these language differences. Like the Afrikaans-speakers, 7g per cent of the National Party adherents favour moderate or pronounced legal segregation. The vast majority of the U nited and Progremive Party

ETHMC AKD STRUCTURAX. FACTORS

members, in contrast, favour limited or no segregation (83 and 99 per cent respectively fall in this category). Further evidence of the close relationship between pohtical preference and home language emerges when political preference is controlled for; the relationship between home language and at t i tudes towards

legal segregation is considerably reduced. It is sharply diminished in most groups, but a portion of it remains intact within the Vational Party subsample. This implies that, while political preference is an i n t ervening v a r iable i n

t h e r e l a t ionship b e-

tween home language and attitudes towards segregation, home language also has a direct dfect on the degree of rigid legal TA B L E

9+

Inkx of Legal Segreganonby Horne Language Home Language

Degree of Segregation Favoured

A&ikaans N=

Moderate/Pronounced Limited/None ..

358

75,5 24,3

Total

Afr.— Eng. N = 95

N = 393

40,0 60,0

I 1,7 88,3

100,0

100,0

Zero-Order Relationship — KcndaH's Tau = 0,630 (p 0,001) T A B t. E

10

laker of Legal Segregation by Political Party Preference Political Preference ~ of Segregation Favoured

National Party N=

Moderate /Pronounced .. * Lim i ted/None

390

72,6 27,4 100,0

United Party N = 174

16,7 83,3

Progressive 4o Party Preference N = 133

N=

102

0,8 99,2

18,6 81,4

100,0

100,0

Kendall's Tau = 0,611 (p ( 0,001) If exclude"No Preference" group, Tau = 0,649 (p < 0,001) +The tables presented in this chapter emerged as part of the analysis of questions dealing with South Africa's future. As a result, those respondents not answering tbe questions about the future have been eliminated from the sample, thereby reducing the nmnber froxn 925 to 888. Interviewees not answering the specihc questions being considered were also eliminated, further shrinhing the base of each table by a s m all amount.

ET%ABC AND STRUCTt1ULL FACTORS

segregation favoured. This would account for the wide spectrum of opinion with regard to segregation found within the National Party. Both ethnic background and religious preference are related to the segregation index in much the same way as home language. About yo per cent o f t h e r espondents who ar e A f r i kaners by

birth, or who belong to the Dutch Reformed Church. favour moderate to pronounced legal segregation. In c o ntrast, over 8g

per cent of the respondents belonging to other religious or ethnic groups favour, at m ost, limited segregation between the races. Both of the v a r iables do not h ave a strong d irect eHect on attitudes towards however. In th e case of ethnic

segreg ation,

background, the primary impact is ~ia home language, as holding the latter constant illustrates. Controlling for home language sharply reduce the original relationship between ethnic group a nd the segregation index — it tends to g o t o z ero i n a l l group.. The same occurs with religious preference if political preference is held c onstant. Th e r e lationship of r e ligious preference with attitudes towards segregation tends to go to zero, though a portion of it r emains intact in some political groups. It

would seem, therefore, that religious preference influences the dependent variable through its impact on political preference, while e thnic background aHects attitudes towards segregation via i t s

impact on both political preference and home language. Religious preference also has a weak direct impact on the segregation index although it is minimal compared to that of home language and political preference. The interrelationships of the ethnically-linked variables and attitudes towards segregation are depicted below. It wi ll be noted that the arrangement of the background variables re-confirms the causal sequence outlined in chapter a. The ethnic cleavages in elite attitudes towards racial segregation are similar t o t h ose outlined by o t her r esearchers dealing w ith d i H erent p o p u l a tion s egments. O v e r t h i r t y y e ar s a g o

MacCrone (t9q7 ) noted similar signi6cant diHerences among Afrikaans- and English-speaking university students in their attitudes towards Africans. He found language group membership to be the most important f actor aHecting at titudes. More con-

temporary research by Pettigrew ( t958 ) at the Univetsity of Natal revealed that the racial attitudes of Afrikaner and Engl ish subjects diHered c onsiderably. S ubsequent w or k b y Pettigrew (i96o), as well as that by Lever (t968) and Rodgers and Frantz (>96z), indicates that language group membership is the most important correlate of the degree of social distance felt

66

ETHNIC AND STRUCTURAL FACTORS

The Interrelationship of the Ethnically Linked Variables and Attitude Variables Ethnic Background Home Language Religious Preference

Political Preference

Attitudes

DesirabiTity of Legal Segregation (Tolerance of Dissent) towards Non-Whites. Afrik«tans-speakers tend to be more reserved in t h eir a t titudes towards Africans, Coloureds and Indians than is the case for English-speakers. This is in keeping with the 6ndings just outlined for the elite sample, which indicate that Afrikaners favour a greater degree of separation between the races than do the English. Despite their high sociw economic position, therefore, elite member' attitudes appear to be much like theme found in other segments of the White population.

Some of the other background variables not discussed thus far (e.g., income level, Non-White contact or travel abroad) are relatively strongly related to the segregation index, but do not 6t into the causal sequence outlined above, Income level and degree of Non-White contact tend to have a negative impact on segregation index scores. That is, as yearly income and interaction with other racial groups increase, the proportion of respondents favouring limited or n o l e ~ ~ egat i o n i n creases. Travel abroad has a similar negative relationship with the segregation index in that as travel becomes more frequent, less legal segregation is favoured. Of the respondents who have never been abroad, 7o per cent favour moderate or pronounced segregation, while among those who travel more often no more than a third favour such a high degree of legal separation. Sixtythrm per cent of those going abroad every few years (seldom) favour limited or no legal segregation and so do p8 per cent

ETHNIC AND STRUCTURAL FACTORS

of the frequent traveGers. This implies that travel, and exposure to foreign environments, mitigates the belief in r i gid legal segregation between the races. T HE IMPACT OF OCCU P ATIONAL SE C TOR ON ATTITUDE S

I n m os t c o u ntries i n dustriaKsation an d u r b a n isation h a ve tended to promote structural or institutional differentiation. Cer-

tainly within South Africa the rapid urbanisation of the recent past has had an impact on th e solidarity of A f r i k aner national-

ism and has led to the development of new bases of differentiation (in this respect see Welsh, Ig69: R75). Some insight into the relative impact of these newly emerging structural factors on individual attitudes can be gained by examining the relationship between occupational sector and beliefs about the desirability of legal segregation. Within the South African context it is also important to contrast the eHect of sector on these atti-

tudes with that of the more traditional ethnic variables, home language and political preference. The original relationship of occupational sector and attitudes towards segregation is a moderate one, as table I I i n dicates. Within most sectors the majority of respondents favour limited legal segregation at most. The only exception to the general trend is within the polity where the majority of respondents favour moderate or pronounced separation between the races. Clearly the relationship of sector and the segregation index is c onsiderably weaker than t h a t d i scussed earlier fo r h o m e language (Tau = o ,63o) or political preference (V = o,6zy). In keeping with this, the two ethnic variables are not strongly related to occupational sector. In terms of home language, the polity is largely Afrikaans, with the exception of local government which is largely English.+ The sectors of mass media, national organisations, professions and economy are abo largely English. The remaining two sectors, religion and education, are fairly evenly split in terms of language group Inembership. The political composition of the various sectors is somewhat more diverse than is true of the language breakdown, as table Ia indicates. Almost two-thirds of the respondents in the pohty prefer the National Party (again, local government forms an egaMost large cities in South Africa are populated predominantly by Enghsh-speakers. Consequently, the city councils in theseareas are dominated by English representatives who favour the opposition parties. Thus, this geographic consteBation accounts for the exceptional status of the local government elite within the polity. It a ho allows local oKcials to exert pressures which counteract the power concentration in the central government. (See Charton, st aL, 1970).

ETHNIC AND STRUCTt:RAL FACTORS T ASt. E

I l

InCkx foLegal Segrego4iorc bp Occspotiortof Sector Degree of Segregation Favoured Sector

Moderate/ Pronounced Politys ..

Education and Research

Limited/ None 40,4

34,5

Religion

65,5

Total

(No. of Cases) 100,0

(359) 100,0

(»)

63,6

100,0

(55)

National Organisations

35,0

65,0

100,0

Kconotny

29,8

70,2

100,0

Mass Media

18,2

81,8

Professions

(60)

(171) 100,0

(66)

100,0

(48) Cramer's V = 0,322 oThis sector includes a number of occupational groupings considered separate sectors in other chapters: civil service, legislature, judiciary and local government. This consolidation of the political sectors is maintained throughout this chapter.

ception as the United Party is dominant in this sub-sector), but otherwise the proportion of National Party adherents is fairly uniform within the various sectors ranging from gee to 4g per cent. The lowest proportion of government supporters is found within the economic sector, but the Progressive Party does not enjoy pronounced supporteither. Rather the economy has the largest proportion of United Party members (R7 per cent) of any sector, as weH as the largest group with no party preference (r9 per cent). The more liberal Progressive Party makes up about a third or more of four sectors: religion, education and research, mass media and professions. In the last two sectors especially, the Progressive segment is large enough to chal-

lenge the conservative hegemony of the Nationalists. These slight relationships of sector with the two ethnically linked variables tend to belie the true significance of occupational context as a causal factor. This comes to light only when sector is linked to the relationship between attitudes towards segregation and home language or political preference. ControHing for occupational sector tends to reduce the original strong relation-

KYKVHC AND STRUCTURAL FACTORS TAnt.E

12

Po(iricul Putly Prcfcrcncc by Occuputionul Actor

Occup taional Sector Education Research

Political Party Preference National

United

62,1

23,1

No Pref.

l1,3

100,0

9,1

100,0

333

13,0

100,0 (54)

18,0

11,5

100,0 (61)

27,1

20,0

18,8

100,0 (170)

14,8

36,1

13,1

100,0 (61)

and

20,5

National Or ganisations..

45,9

Economy..

34,1

Professions

Total

(No. of Cases)

273

17,0

(346) (88)

100,0

(47) Cramer's V = 0 +14 ship between home language and the segregarion index somewhat, as table 13 shows. The relationship remains most pronounced in those two sectors where the language split is relatively even; that is, half of the respondents in the sectors of rehgion and education /research speak A f rikaans, while h alf speak English. These two sectors are also those within which organisations are exclusively uniligual. In contrast, within the sectors of mass media and professions there is considerable interaction between language groups and it is in these that the rdationship between home language and the segregation index is most sharply reduced. The majority of professionals and journalists favour limited segregation regardless of home language.

This means that the Af rikaans-speakers in these two sectors are

more likely to favour limited or no legal separation than are Afrikaners in any other sector, perhaps because the proportion of English-speakers and, consequently, liberal attitudes is more pervasive than is generally the case. Similarly, the Englishspeakers in the sector with the highest proportion of Afrikaners — the lrohty — are less likely to favour bmited or no segregation than are the English in other sectors. In this case the pervasiveness of conservative attitudes seems to m i t i gate E n g lish

opposition to legal segregation somewhat.

$0

E~

C AND S T R UCTURAL FACTDRS

va sx.a 13 Proportion op Rrs~

F ao onring Litnitet or no Legal Srgrogation bp Horns Longnags antf Oornpational Mtor Degree of Assoc.

Home Language

Occupational Sector

A&ikaans N

A&.— Eng. N ~/

(Ken dali"s Tau)

102 77,5

209 20,6

0,506

(p < 0,001)

Education and Research

39 28,2

8 1000

41 95,1

Religion

29 31,0

0

26 100,0

0,0

0,660 (p < 0,001) 0,688

(p < 0,001) ; 4ational Organi rations ..

Professions

18 22P

9 6 6,7

32 87,5

35 14,3

22 63,6

19 52,6

2 5 0,0

45 95,6

6 83,3

33 93,9

9

5 5 ,6

358 24,3

0,584

(p < 0,001)

95 60,0

0,543 (p < 0,001) 0,374

(p < 0,001) 0,260 (p < 0,01) 0,630

(p < 0,001)

Controlling for occupational sector has a similar dfect on the relationship between the segregation index and pohtical preference as it has on the relationship between home language and attitudes towards segregation. Although the impact of political preference on segregation attitudes remains fairly clearcut in all sectors, as table .ty indicates, the degree of relationship varies. (See the statistics at the right side of the table.) The sectary in which the original relationship is strongest and weakest are the same as those referred to in the previous discussion of home The impact of pohtical preference is most pronounced in the sectors of religion and education/research.+ In

a ge .

«The relationship between yolitical pxeference and the segregation index is also qmte pronounced in the nadonal organisations sector. The reasons for this are presumably sixnilar to those accounting for the strong relationship within the religious and educational sectoxs. Civic organisations, like universities and churches, are often oriented to only one language or yolitical gxouy and, consequently, the members of any organisation are cut oK froxa mexnbers of other groups. As a result, attitudinal differences between political groups remain pronounced within the national organisations sector.

ETHNIC AWD STRUCTURAL FACTORS TASL E

14

Proportioa of llcspoarfcatr Faoostriag Limital orao Legal Segregation by Political Profcrcaco and ocrupntiorll Sector

Occupational Sector

Political Party Preference National /0

United

;lj

0

/

Progres. o/o

Xo Ptef.

Degree of Assoc. (Crarner's

V)

N

207 20,3

79 81,0

12 160,0

35 62,9

0,580

Education and Research

37 27,0

17 82,4

24 100,0

18 87,5

0,681

Rehgion

27 29,6

2 100,0

18 160,0

6 100,0

26 26P

15 100,0

11 100,0

7

Kconotny

54 29,6

45 80,0

33 97,0

Mass Media ..

20 50,0

9

Professions

18 72P

8 106,0

390 27,4

174 83,3

Polity..

National Orgamsatlons

Zero-order Relationship ..

7 'l,8

8 5,7

6,740

31 93,5

0,621

22 100,0

8 100,0

0,561

13 100>0

7

7 1,4 0,627

both of these sectors the proportion of Progressive Party adherents is fairly large but they are clustered in the Englishspeaking organisations. Similarly, the Afrikaners or National Party supportets stick to their own universities or churches and do not expose themselves to English inHuences. Consequently, the divergence in attitude between the political groups remains pronounced in these two sectoss. In the other two sectols where the Progressive segment is sizeable, in contrast, the relationship between political preference and attitudes towards segregation is weakest. In the professional sector the relationship is reduced by one-third, while in the mass media it is retluced by Io per cent. In both of these sectors the large proportion of National Party respondents favouring limited or no segregation (go to ttg per cent) accounts for the reduction of the relationship. Both professionals and journalists must interact quite frequently with members of other language and political groups and this may account for the liberalisation of N ationalist attitudes, as the ruling party accounts for a lesser proportion of the sectors of professions and mass media (35 per cent approximately) than is the case in other sectors. In sum, although occupational sector is relatively weakly related to the ethnicaHy-linked background variables, and only moderately related to attitudes towards segregation, it is none-

72

ETKBAC AND STRUCTURAL FACTORS

theless important, as a conditional factor in some instances. In part the association between sector and the segregation index is, of course, due to the language and political composition of occupational groups. Thus, the largely Afrikaans and Na,tionalist sector of the policy has the highest proportion of respondents f avouring moderate or pronounced legal ~ atio n (see table t r). Simihrly, in the English-dominated economic sector almost three-quarters of the respondents favour limited separation bet~veen the races. This tension between the political and econo-

mic sectors of South African society has been discussed in some detail by numerous writers (see, for example, D. H o bart Houghton, >96y). Although the attitudinal differences between civil servants and businessmen are largely due to their divergent political and language affiliations, sector does have an impact on the segregation attitudes of a number of other occupational groups. Vy'ithin the sectors of professions and mass media factors within the work environment serve to mitigate the impact that

horne language and political preference have on segregation attitudes within the rest of the sample. As tables t3 and t4 show, the original relationships are most sharply reduced in the professional sector and somewhat less for the mass media. In both cases the reduction occurs because the Afrikaans and National

Party respondents are unusually liberal and half or more favour limited segregation. This liberalisation of attitudes seems to occur for a number of reasons. First of all, each sector contains a sizeable proportion of respondents opposed to widespread legal segregation (68 per cent English-speaking and go-S5 per cent Progressive interviewees in each) , so that a libel i n gi n fluence of consequence is present. Second, there is considerable interaction between the major language and political groups in the two sectors. For, journalists must cover the news regardless of what group it concerns and since everyone reports on the same events, exposure to divergent viewpoints is unavoid, ble. Similarly, professionals cannot pick and choose their cl i ents or patients and must put up with alien opinions as part of their daily routine. Finally, members of both the mass media and professions are part of an international community which me-.ns they encounter foreign liberalism more frequently than dn most other South Africans. Such exposure, along with the pr~~nce of frequently encountered domestic liberalism, serves to undermine the traditionaHy conservative views of the A fri-

'aan.- and National Party groups. The key to such liberalisation of attitudes towards segrega-

ETHI C

M V D S T RUC~

FACTO RS

73

tion lies not so much in the presence of Progressive Party or E nglish orientations, as i n t h e f r e quent i n teraction w i t h divergent viewpoints which is an integral part of the work world of professionals and journalists. The sectors of religion and education/research serve to make this clear. Both of these sectors have a sizeable English-speaking segment (approximately yo to go per cent) and about a third of the respondents favour the Progressive Party. In spite of this hberal component the language and political divergences in terms of attitudes towards

segregation are more pronounced in the religious and educational sectors than in most others. The explanation for this lies in the language and political isolation that characterises the two sector within the South African context. Although language and political divisions occur and coincide in other sectors such as the civil service and civic affairs, these divisions run deeper in the cases of education and religion because of the institutional separanon and contrasting ideological commitment of the two language groups. Unlike the civil service, which is a bilingual institution (although most bureaucrats are uniTingual at home), any given church or university can provide relative isolation and an exclusively uniTingual environment. Such a situation serves to create an atmosphere that re-enforces traditional attitudes and beliefs, making them more pronounced than they otherwise might be. %without interaction with other groups, then, attitude change does not occur.

The exacerbating influence of occupational context within the sectors of religion and education is only a limited one, of course, as is sector's Viberalising impact in the case of professions and mass media. By and large home language and political preference are far more important in shaping attitude towards segregation than is occupational sector. It is weH to remember, however, that South Africa is still in the midst of its industrialisation process and structural differentiation is likely to increase considerably in the yean to come. As sectors become more and more specialised the worlds of various occupational groups will diverge more sharply than is presently the case. Members of v arious sectors x~dl then share a r elatively exclusive frame o f

reference and will hand they have more in common with each other than with their language or political compatriot. As a result, the impact of occupational sector on attitudes may increase sharply and eventually override impact of language or political factors. Such a devejopment would make South Africa more like the highly i ndustrialised countries of present-day Europe and America.

74

ETHNIC AND STRUCTURAL FACTORS

TOLERANcE oF DIssENT (see also chapteI' 8 )

In addition to attitudes towards legal segregation, respondents were questioned about the degree of dissent they deemed desirable. The question relating to this issue was as foHows: "In your opinion should radical leftists or radical rjghtjsts be allowed to do the foHowing: (g) make speeches at publtc Ineetings P

(b) belong to anyorganisation' (c) teach in universities'3" The nature of this question was such that it could be used. as the basis for a Gu~ sca l e, just as attitudes towards legal segregation were. Originally the scale was constructed using replies pertajIung to both radical leftists and radical rightists. This revealed, however, that if a respondent was wiHing to let a radi cal ieftjst dp one of the above things, he was also willing to let a radical rightist do it. I n a ddition, respondents tended tp approve of any one of the three activities for radical rightists before approving of it for radjM leftists. This would seem to be a reflection of the fact that the radical right is viewed with greater sympathy in South Africa than is the radical left. Npt differential tolerance is any too pronounced, however. For, while respondents are more likely to let rightists make speeches atpublic meetings than leftists, they will agree tp radical leftists bdonging to any organisation before they wHl permit radical rightists to do so. That is, respondents tend to djs. tribute privileges alternately rather than approving of aH three activities for radical rightists before granting any privdeges to the radical leftist. In view of this, the scale used for coding this q uestion pertained only to t h e radical leftists, sjnce aHpwed one to assess the relevant variation most simply. (Fpr a more detailed discussion of the scale's construction, see Huber, Ig73: Appendix 3.) The djssent, scale has a range of o to 3, and individual scores are arrived at by assigning a weight of I t o each sphere m which the original question is answered positively. If, however, a respondent states that a radical leftist should not be allowed tp engage in a certain activity, he is assigned a weight of o fpI that sphere. Final scores consist of the total weights assigned to each of the three sectors. The distribution of scores indicates that if respondents grant any rights to radical leftists (i.e., they had a score of I ), they were mast likely to favour their making public speeches. Similarly, the right to teach at unjveIsjtjes js extended only if the right to belong to organisations and to

E~

c A N Ds TRUGTURAL FAcroRs

make public speeches has also been extended. Respondents in this last group encompassed about half of the sample and, con~u ently, the distribution of scores was dichotomised for the purposes of analysis. This meant that those respondents receiving scores of g, and making up @6 per cent of the whole, are considered to have a "moderate" tolerance of dissent. The rem~ 54 p n . cent of the sample makes up the group having a "limited" tolerance of dissent (i.e., received scores of o, t and

~).

The ~ t ind e x i s quite closely related to the segregation index as table t5 indicates. The two variables exhibit a strong, and inverse, relationship in that as the degree of legal segregation increases, the degree of dissent domed proper decreases sharply. That is, g7 per cent of the elite respondents favour limited segregation and ar e m oderately tolerant of d issent, whereas another ga per cent have a limited tolerance of dissent and favour moderate or pronounced legal segregation. Or, to put it somewhat diHerently, 6a per cent of those favouring limited segregation are moderately tolerant of dissent, while of those advocating moderate topronounced segregation, 79 per cent exhibit. a limited tolerance of dissent. T A B 1. E

15

RcfuticnshiP oJ Legul SsgrcgaAonun' Dissent Degree of Segregation Favoured

Tolerance of Dissent

Total

Kendall's Tau

=

Moderate/ Pronounced

Limited/ None

21,0 79,0

62,4 37,6

100,0

100,0

(M4)

(474)

— 0,40$ (p < 0,001)

This association between the two attitude indices is in keeping with the language and political cleavages among the elite respondents. Approximately 6o per cent of the Afrikanets, or National Party manbets, favour moderate to pronounced legal segregation and have a limited tolerance of diment. Englishspeakers and members of the opposition parties, in contrast, are much more likely to advocate limited or no segregation and

76

ETHMC AND STRUCTURAL FACTORS

to exhibit a moderate tolerance of dissent. This similarity between the two in dices leads one to suspect that they have simi-

lar causal antecedents. This supposition was confirmed by a more detailed data analysis which related the background variables to the dissent index (it is not discussed in greater detail because the results are so similar to the analysis for the segregation index presented earlier). It ap peais that political preference has th e s t rongest direct i m p ac t o n

a t t i t u des t owards

dissent, while ethnic background, home language and religious preference have largely an indirect impact through their effect on other background variables. Home language and religious preference aho have a limited direct impact on the dependent variable. Consequently, it seems fair to conclude that the causal sequence outhned for the segregation index aho applies to the index of dissent. Of course, since the relationship between the two dependent variables is an inverse one, the im-

pact of the ethnically linked social characteristics on tolerance of dissent is the opposite of that isolated for attitudes towards legal segregation.

This chapter analysed the impact of t raditional language d ivisions and

r a p i dl y e xpanding s t r u ctural d i f f erentiation o n

several attitudes. The Afrikaans- and English-speaking elites vere compared in terms of their attitudes on two crucial issues: racial segregation (currently known as petty apartheid ) and tolerance of dissent. Analysis of th e questions dealing with these two topics revealed that the majority of the elite respondents have a

l i m i ted t olerance of d i ssent and f a v ou r l i m i t ed

segregation. In both cases the majority is not great, however, because yt per cent of the sample believe there ought to be moderate or pronounced legalsegregation and 46 per cent exhibit a moderate tolerance of dissent. When th e tw o a t t i t ude variables are related to e ach o t h er

they exhibit a fairly strong inverse relationship which is in keeping with the ethnic and political cleavages within the sample. That is, A f r i k aans-speakers and R a t i onal P arty m e m bers are more likely to f avour moderate or p r onounced segregation and

have a l i m i ted t olerance of d i ssent. English-speakers and members of the opposition parties (United and Progressive), in contrast, tend to believe there ought to be limited legal segregation and exhibit a moderate tolerance of dissent. In a d dition,

the other two ethnically linked variables, ethnic background and religious preference, have a share in shaping the two atti-

ETHNlC AND STRUCTURAL FACTORS

77

tudes under discussion although their impact is l argely indirect. The decisive impact of political preference and home language on attitudes towards l~ se g regation is in keeping with other msearch dealing with a w i der spectrum of South Africans than our ehte study did. Previous researchers concluded, as we do, that ethnic identity is the best predictor of attitudes concerning relations between the racial groups.

During the coulse of analysing the impact of various background variables on segregation attitudes, it emerged that occupational sector serves as a conditional factor in the relationship between home language or political preference and the segregation index. That is, i n s ome instances occupational context serves to i n crease or decrease the original relationship

between attitudes and background variables. Within the sectors of professions and mass media the differentiating impact of home language and political preference on attitudes towards segregation is considerably reduced because the Afrikaners in these groups are more liberal than is generally the case. It was argued that the Afrikanets' unusual preference for limited segregation is due to the nature of journalistic and professional pursuits which make daily interaction with outside viewpoints (e.g., English, Progressive Party or foreign) an inevitable part of daily life. What makes the sectors of professions and mass media different from the others is not the mere presence of liberal viewpoints, but having to cope with diversity constantly. In sectors where different language and political groups exist in relatively large nu mbels, but i n

i s o lation, tr aditional a t t i t udes

do not change. The religious and educational sectors illustrate this tendency because the various institutions within them exist in language and political isolation. Consequently, the of the ethnicaHy linked variables on attitudes towards segre-

impact

gation is somewhat greater in these sectors than is the case for the rest of the elites. A t present the i m pact of sector o n a t t i tudes is li m i ted a n d

does not challenge the dominance of ethnic influences. By and large the sectors which are most strongly dominated by Afrikaners and National Party adherents (e.g., polity) have the lowest proportion of respondents favouring limited segregation, wMe the highest proportion of respondents desiring legal separation is found within the sectors where English and opposition party representation is the strongest (e.g., economy and mass media}. Nonetheless, as South African industrialisation continues in the years to come, structural differentiation will become more complex than it is at present. As sectors become

y8

ETHNIC AND STRUGTURAI FACTDRS

occupa tional

more specialised, the impact of c ontext on a t t t tudes will no doubt increase and eventually challenge the pre sent dominance of ethnicaiiy linked factors. The next chapter focuses in greater detail on another dimension of structual differentiation; we turn our attention from occupational sector to social class divisions in %hite society.

CLASS DEMVATIOJN AED I'ERCEP1 IOUS OI' AIMEAAES- AED EJfGLISH-SI'EASING ELI1 ES iKTRODU CTIOK

This chapter analyses the class derivations of the elites and their perceptions of the South African class structure. The socio-economic proSe of t h e d i tes presented in chapter shows to what extent they constitute a highly select group in South African society. Class derivations of the eiites and their perceptions of the class structure should be analysed against this background. While the elites do not represent the huger society, knowledge about them and their views contributes greatly toward understanding trends in t h e

l a rger society. Social stratiTication

patterns of whole societies cannot be studied without proper analysis of elite groups whose stratification is related to the general sorel structure. The internal stiatification patterns of elite groups have a significant impact on the larger society and is thus "a theoretical prerequisite for a m ore comprehensive understanding of modern social structure". (Janowitz, i 956:

8.)

CLASS DEiuvATiONS AND PRESEhT STATUS

Historically, Afrikaans-speakers have been at a disadvantage in relation to Engbsh-speakers in the political and economic fields in South Africa. During the past decades, however, they have made great gains in the political fleld and have also improved their economic position.

Anal)~ o f

th e A f rikaner work force in the cities by Pauw (rgb) and Van Wyk (i968) have r~ r d ed in great detail the influx of Afrikaners into urban occupations. While they have entered at all levels, the percentage increase has been much greater at the lower status levels. (See Van Wyk, i968, Tables 8.i, 8.a, 8.3.) The general conclusions of Van Wyk are that the Afrikaners in i96o were at a considerable disadvantage in the White occupational structure of South Africa ( i968: zz i ),

8o

CLASS DERIVATIONS AND PERCEPTIONS

since they were concentrated in occupations to which relatively smaH economic and other benefitswere attached. Comparatively few Afrikaners were in the professional and administr tive categories or had Isnerged as entrepreneurs and employers. Afrikaners were in the minority in the private sector but were well represented in the pubhc service of the state, the provinces and the cities. In I966 it was stated (Van Kyk, I968: Ram) that in the civil service p I per cent of the Wh i te workers were Afrikaans-speak-

ing and 29 per cent English-speaking; in the police 87 per cent were Afrikaans and I3 per cent English; in the prison service 98,g per cent were Afrikaans and I,y per cent were Englishspeaking. Afrikaners came to dominate the public service in South Africa. Van AVyk (I968: 228-rag) cites several figures to iHustrate the great economic gap between Afrikaans- and English-speakets in terms of peIsonal income and control over the economy. In Ig3y-qg the average annual income of Afrikaners was Raga while that of Whites speaking other languages was Rrg6, i.e. a ratio of I o o t o I 2 8 . T hi s ratio i n I g g g was Io o t o I 7 o , a n d m 1964 It was loo to I 30

The Afrikaner has also entered and extended his control over larger areas of the commercial and industrial fields. Between I939 and I964 Afrikaners increased their share in trade (in terms of turnover ) from 8 to 28 per cent; in industry (in terms of production) frotn 3 to Io per cent; in finance (in terms of monetary control) from g to I4 per cent, and in Inining (in terms of production) from I to I o p er cent. While these gains are considerable, Afrikaners were still under-represented in the early sixties in view of the fact that at that time they constituted g8 per cent of the White population. AVhHe their standard of education during the past decades has improved markedly, Van Wy k de monstrates that Af r i kaners were, by the I 9 6os, stiH lagging behind the Englishspeakers in this respect. (I968: z 3 I.) Although twice as many Afrikaans- as English-speaking children entered school, so many of them d r opped out that an equal number of b oth l a n-

guage groups completed high school. Furthermore, of t hose who completed h ig h s chool, r e l atively f ewer A f r i kaans students proceeded to university. In I96I, g7 per cent of Afrikaans students and 46 per cent of English students who matriculated enrolled a t u n i versities. I n I 9 6 o , y z o oo E n g lish-speakers obtained degrees and diplomas in South Africa as against 37 tioo Afrikaans-speakers.

CLASS DE1UVATKONS ABD PERCEPTIONS

Sx

In i963, only a7,4 per cent of all engineering students and 33,6 per cent of all medical students at South African univers ities were Afrikaans-speaking, while g8,6 per cent o f l a w students were Afrikaans-speaking. (Van Wyk, i968: a3a.) The composition of th e elites interviewed should be seen against the background described above. In the analysis below, we will compare Af rikaans- and English-speaking elites in terms of occupational status, education an d

i n c ome.

Occupational status

For the purpose of analysing the occupational status of elites and their fathers, we classi6ed the present occupations of elites, their initial occupations, and th e o ccupations of t h eir f a t h eis

at the time respondents left school into the following 5 occupational status levels: i. Top Level: Th ese were persons of authority, holding the highest positions in unportant structures, organisations and companies. They included members of th e cabinet, judges, provincial administrators, secretaries of state departments, principles of universities, and directois and managers of commercial and industrial companies with a s sets of o ver R 8 o ooo, a nd o f f i n ancial

concerns (banks, e t c.), wi t h as s ets of over Raoo ooo ooo. r. Higher Pro fessional and Executive: These were men in professions wit h

i n c o mes over R i o o oo ; b u s inessmen

earning over Rao ooo; those holding top p(i t ious in all large companies that did not qualify for the top level; editors and managers of )Vhite daily newspapers with circulations exceeding 6o ooo; town clerks of cities with populations o f

o v e r i oo o o o ; d e p u t y s e cretaries an d

higher ranks in the civil service; inspectors of schools; professors; members of parliament, etc. 3. Lou,er Professional and Executive; This included professionals earning below Rio ooo; businessmen earning between Rio ooo and IBo ooo; local government officials; teachers and university lecturers; church ministers; com-

missioned police; civil servants below deputy secretaries, etc.

4. Clerical, Service, Trade and Skilled Work: This included n on-commissioned pohce men, messengers of c o u r t , clerks, technical an istants, shop keepers and skilled workeis.

cLAss DKiuvAYtoivs AÃD pERGEprtows

g. Semi of the sample did not reply to this section of the question, and replies cannot be considered typical of the whole group. Of those who did reply, however, the great majority illentified the radical left chiefly w ith authoritarianism and totahtarianisln — ao per cent o f

the

whole sample. In defining the radical right, the following components were recognised: (a) Fascist, nazi, Totalitarian, Authoritarian. (b) Racialist, Herrenvolkist. (c) Ultra conservative, bigot, reactionary, not prepared to change. (d) '%'erkrampte", extreme nationahst, exclusive group consciousness, ideologist. Once more a largepercentage (@9 per cent) did not respond, and there was a further group of r esponses not possible to classify in t~ of t h e above criteria — Ig per cent. In this case the image of those who responded was of an ultra-conservative (Ip per cent of the total sample) or of a "Verkrampte" (9 per cent). The most prevalent image of a radical leftist was thus of a totalitarian. The radical rightists, on the other band, were chiefly identified with ultra-conservatism and nationalismnot with racialism and totalitarianism or authoritarianism. It

is tempting to attribute the relative tolerance toward the right to the difference beevtxxl the images of left and right. It is also tempting to speculate on the psychological reasons which might lie behind these different images. The White elite exaggerates the fear of the left because of its political connotations; the radical right which stands for the survival of the W h ite

i45

SOME POLITICAL ATTrf UDES OF THE ELITES

group is seen as much less of a bogey because it does not threaten his position in society as does the radical left. Political i ntermarri age A further question attempted to chart reactions to political intermarriage by asking: "Provided the person is acceptable in other respxts, how would you feel if your son or daughter wished to marry into a political group different from your ownÃ' Five distinct political orientations were listed, i.e. Radical Left, etc.

It is evident that elites tended to disapprove of intermarriage of a political nature to a g reater extent than o f a l l owing radi-

cals certain basic pohtical freedoms. While only i4. to R6 per cent of all elites deny the radical Right and Left basic political freedoms respectively, the corresponding percentages disapproving of intermarriage are i) and 4o per cent. While ~ per cent of all elites allow radicals basic freedoms, only aa per cent approve of intermarriage. The repulsion towards those on the radical left is then exaggerated at personal leveL This could indicate not only psychological concern for the survival of the White group, but also a degree of fear for the personal consequences of such a marriage in the existing political situation where radical opinion on the left has been, politically spealang, outlawed, but not that on the radical right. Correlation with political party preference, language group, political participation and religious ~ t ion f o l low the same patterns as those revealed in the question with regard to the rights of radicals. E VA L U ATION OF C E R T AIN INTEREST A"4D PRE SSL~

GR OU P S

Because of the ethnic polarisation in South African society, many structures and voluntary organisations are endowed with an ethnic and culture group identification. For instance, the South African Broadcasting Corporation (SABCI prior to Igg8 was seen as the creation of the United Party government, and therefore as being iniinical to Nationalist interests. In the postIgy8 era the identification has been reversed, with the change of government, and the SABC is now seen as the creature of the Nationalist government, and as inimical to English interests. The whole structure of group interests in South Af r ica tends

to be in conformity with the prevailing ethnic cleavage. Thus there is a Red Cross Society, predominantly English-speaking,

I $6

S OM E PO LI T ICAL A~E

S O F TH E E L ITES

and a N o odltulpliga, predominantly A f r i kaans-sp~ . A number of these groups were selected with the object of testing reactions for or against them. We were interested in determining attitudes toward some interest groups and pressure groups which have become relatively controversial because of their identification with either language or political groups. In the question respondents were asked whether the group listed exercised an a d v antageous, mixed o r

d e t r imeIItal in flu-

ence on South Africansociety.The following groups were bsted: Two secret societies, the Afrikaanse Broederbond (AB) and the English-oriented Free Masons (Flail); an Afrikaans cultural organisation, the Federasie van Afrikaanse Kultuurverenigings (FAK) and an English-oriented South African service organisation, Rotary (Rot.); the Afrika~ rie n ted South African Broadcasting Corporation (SABC) and the liberal interracial Christian Institute (CI); the English and Afrikaans press and churches; the Chamber of Mines (CM) with a dominant Enghsh-speaking economic elite and the Trade Union Council of South Africa (TUCSA), predommantly Englishspe~ , m o u thpiece of White, Coloured and Indian workers. There is a polarisation between the two major parties. The opinions of Progressives and Liberals follow the pattern of the United Party supporteIS and were not included in Table 26. In a number of cases large percentages did not answer the question. This was especially true for the two seel' societies about which not much is known to the public. But it is also TABL E

26

Imager of Iofhtlc of CertainIrttcrctt Groups Itctotcdto Party Prcfcrlcc (Pcrccntogcc)

Group or Institution

N.P

A.B. F .M.. . S.A3LC. C.I. A. Press

C. of M. T.U.C.S,A...

26

0 31 74 13 79 72 12 6 40 22 7 16 23 64 36 29 43 51 79 12 43

All

Elites N.P

V.P

13 16 39 49 38 23 18 16 45 35 59 26

19 22 17 37 24 21 70 68 29 55 28 44

6 32 41 1S 70 26 76 72 54 51 13 45

15 25 27 24 46

22 70 67 42 50 23

15 25 6 0 48 2 18 26 6 17

79 7 41 17 14 1 7 7 3 14 23 6 4 3 0 2 6 4

31 9 4

10

37 15 44 7 29 32 7 25 2 2 23 28 2 8 5 0 2 5 2 15 24

9 8

29 40 20 17 3 26 33 7 6 13 18

SOME POLITICAL ATTITUDES OF THE ELITES

I47

obvious that the reason for not answering the question is not merely lack of knowledge, for the Free Masons (about which yo per cent did not answer ) is much better known and less secretive than the Broederbond (about which Eg per cent did not answer). Furthermore, a much larger percentage of Nationalists, who are obviously better informed than United Party supporters, did not answer the question on the Broederbond. Md almost twice as great a percentage of United Party supporteIE refused to answer the question on the Free Masons, an or~ tion a b out which they are obviously better informed than the Broederbond. In a n ormal society where there is a reasonable measure of consensus on basic values and means of achieving them, one would expect the major organisations and institutions to be rated as having either advantageous or mixed influence. To be rated detrimental is an indication of severe divisions within that society. This eagerness to express an opinion is a function of fear and resentment and not of knowledge. It is motivated by the same emotions that cause elites to over-estimate the power of their enemies as described in chapter g. In our list of prominent economic, political, cultural, religious and service organisations, only one economic organisation, the Chamber of Mines, was rated by more than 5o per cent as an advantageous influence (gg per cent ). Other groups which received an advantageous rating from more than one-third of the elites were Rotary (49 per cent), Afrikaans Churches (45 per cent), the F.A.K. (89 per cent), the S.A.B.C. (38 per cent), and Engbsh Churches (8g per cent). Only one group received an u nusually high d etrimental rating and that is the Broederbond which was rated de~ e n t al by yI per cent against I3 per cent advantageous and I5 per

hang

cent mixed. A f ter the Broederbond, the Christian I n sbtute re-

ceived the highest detrimental rating of c3 per cent, against a rating ofE3 per cent advantageous and sa per cent mixed. The third most detrimental group is the Free Masons (I y per cent ). These three groups which were rated most detrimental are also those on which most people declined to answer. While the Broederbond was regarded with an almost singular lack of confidence by the vast majority of elites, a complete polarisation of opinion between the two parties is evident from the fact that all those who rated it as advantageous are Nation-

alists, and the vast majority of those who rated it as detrimental are United Party supporters. It was the only group which received the unanimous disapproval of the Umted Party. But

148

SO M E P O L1 T JCAL ATT1TUDESOF THE EL1TES

not even among Nationalists did the Broederbond enjoy a very good reputation. Only a6 per cent saw it as advantageous, 19 per cent as mixed, 1g per cent as detri1nental, and 37 per cent refused to answer.

Nationalists were unanimous in their approval of the F.A.K. and the S,A.B.C,, both of which might be seen as having an educative and cultural f u nction in th e c o m m u n i ty, particularly

in the Afrikaans community. Both, however, may be regarded a s having political significance, which w o uld a ccount f o r United Party disapproval of their influence. Very smail numbers of either party disapproved of the Afrikaans press — a party press for the most part — or the Afrikaans churches. Significant numbers of Wationalists disapproved of the Free Masons, a secret society, the E nglish Pr ess, the Trade U n i o n

Council, which was espousing to a significant degree the cause of the Black worker, and the Christian Institute. It is for this latter organisation that they reserved the heaviest weight of d i s -

approval, no less than y8 per cent seeing it as having a detrim ental influence. Only 6 p e r cent, on th e o t her hand, saw th e

English churches as ha~kg a detrimental inHuence. It is possible that the liberal line of the Christian Institute is seen as more of a theat than that of the English churches because of the participation o f in its activities.

s everal o utstanding A f r i k aner theologians

The United Party adherents reserved their most unanimous approval for the Chamber of Mines, and for Rotary, an organisation with objects stressing community service. It is perhaps signi6cant that th e U n i ted Party opted fo r a s u ccessful econom ic institution, and f o r a n e l ite social cl ub . T h e N a t i onah~ , on the other hand, opted for an organisation and an institution b oth of w h i c h h av e a c u l t u ral s i ~ can c e w i t h i n t h ei r o w n

group. Theadvantages of the F.A.K. are rated even higher than those of the Afrikaans churches. In a q u estion l ik e t h i s w h ere respondents have the o p t ion

of rating an organisation as having mixed influence, that is both a dvantageous and d e trimental, but n e vertheless prefer to r a t e it as detrimental onl y, one c an c onclude t hat t he d i s approval o f such an organisation must be i ntense. The f ailure of t h r e e organimtions to obtain even a m i xed r a t ing f rom l ess than 5o

per cent of the elites (Christian Institute y5 per cent; Free Mai n s y t p e r cent ; B r oederbond r8 per cent) is an indication o f their d i v isive f o rce i n S o ut h A f r i ca n s ociety. B y f a r a n d away the group thought to have the most detrimental influence

was the Broederbond.

SOME POLITICAL ATTITUDES OF

THE ELITES

I49

Both parties revealed significant degrees of intolerance of the secret societies, the Broederbond and the Free Masons; the Nationalists also revealed considerable intolerance with regard to the C h ristian I n stitute, a church o r ganisation which

fulQs an overtly political function, and wham approach to the race problem in particular is extremely liberal. The intolerance of the United Party for the Broederbond is clearly not only because it is a secret society, for the F ree M a sons suEer from aiinost the same disadvantage but d o n o t c a l l f o rt h t h e same

unanimous disapproval from the Nation~ . The B r o ederbond, however, is seen as fulfilling a secret and sinister pohtical role and is feared as an instrument of political domination. F rom these results one might infer that m ost U n i ted Party ad -

herents distrust and fear the Broederbond and the militant Afrikaner nationalism which it seeks to promote. This then may be seen as an intolerance based on a fear of ethnic domination. On the other hand, most Nadonalists distrust and fear t he Christian Institute which poses an ideological threat to t h e

Afrikaner Nationalist group. The minority group evinces fears of ethnic dominance; the majority group evinces fears of ideological subversion. CONCLUSIONS

The questions on attitudes were selected at random in order to test for significant differences beuveen language groups. They have certainly revealed that the ethnic cleavage of the White S outh A f r i can p o p u l a tion i s c o n sistent f o r f a r mo r e t h a n

language, field of experience and pa~ tag — at elite level anyway. And a detailed exposition of these basic differences would do a great deal to elucidate the con8ict which has taken place in the past between the two groups.

It is apparent that E nglish- and A f r i kaans-speakers have rather different concepts of democracy. M any A f r i kaners would

appear to stress the role of the group, for what is important to a comparatively large group is the mill of the people. Freedom i n the past has been important to th e A f r i 4 m er , bu t w ha t h e c ame to know as freedom was the freedom of t h e group f r om

alien domination, and the freedom of the individual within a group founded on a common religion and with common values. In contrast to this group centred attitude towards democracy,

the English-speaker appears to lay the chief emphasis on the concept of consent: and here it is the individual who is more important than the group. These differing perceptions of the individual in relation to

I$0

SOME POI,ITICAI. ATTITUDES OF THE EXSZES

the group also come out in replies to the remaining questions. Thus Afrikaners tend to see their political party as an ~ ment for the realisation of group values in many different fields. The English-speakers see their parties as vote catching, policy making, and as recruiting organisations. They see them then as groups of collaborators joined together for rather specific and limited purposes. The Afrikaner, whilst eschewing ideas of state socialism has nevertheless been prepared to accept a m easure of n a tionalisa-

tion, when it could be proved to be in his group interest. English-speakers would appear to prefer the unfettered free enterprise solution. The Afrikaner, while by n o m eans intolerant of p olitical deviation, is nevertheless less tolerant than his English-sp~ counterpart. And here again one might advance the theory that the English-speaker's political tolerance is the amdt of his ideas of individual freedom. Men have a right to different opinions; for the Afrikaans-speaker they have that right just so long as they do not endanger the safety of the group, and they may d iffer only w ithin fhe agreed conceptual framework of t h e group. The polarisation of South African White society is vividly reflected in t h e a t t i tudes towards certain group i n t erests. The

minority group evinces fear of ethnic domination; the majority group feats ideological subversion at a r eligious level — its Achilles heel perhaps.

PRELIMINARY DEFINrrlONS

This chapter explores a number of ideas about the future articulated by various segments of the South African elites.f Since images of the future are quite diverse and encompass a broad range of topics, it was not possible to investigate aH types within the confines of a single research project. Consequently, a number of working definitions were outlined at the start of the research to delimit the paths of inquiry pursued. These definitions referred to three different images of the future: expectations for the future, hopes for the future, and fears for

the future. Each was de6ned as follows: Expectatiortsfor the future are relatively specific ideas about the characteristics of a future state of affairs and include the belief that the projected state wiH someday become a realip. Often statements of strategy, delineating a method for reaching

the projected end-state, are clom,ly associated with expectations.

Hopes for the future are ideational constructs of the best possible future an individual can foresee. They are usually based on the foHowing premise: "if everything proceeds exactly as I w ould like it to, the future wiH be as follows.. . " Fears for the future are statements referring to the negative aspects of the future. They depict the probable state of awaits if "everything goes wrong". The three images of the future vary both in terms of the way they are evaluated and in terms of the probabihty of realisation attached to them. Both hopes and feaIs have positive or

negative evaluative bases as integral parts of their make-up. Whether either wiH become a r eality, however, is not com+For a considerably more detailed discussion of this topic, see Huber, 1973. tThe tables presented in this chapter have a somewhat smaller base than

is true of those in most other chapters. This is because an unusually large number of respondents did not answer most of the questions dealing with the future. Consequently, they were eliminated from consideration reducing the sample size to 888. fn addition, respondents not annvering a specific question were deleted from anal'~ in v olving that query thereby further shrinking the number of respondents somewhat.

IMAGES OF THE FUTUR E

pletely clear as far as the individual propagator is concerned. Expectations for the future, on the other hand, are conceived of

as the most likely future an individual foresees and have the greatest likelihood of realisation of the being discussed. whether a n e x pectation i s p ositive o r n e gative, however, depends on the individual's value structure and is not immediately apparent from the image's content. The goodness or bad-

images

n ess of this future idea is not, therefore, an i n tegral part of i t , as is the case with hopes and fears. Although the various images of the f u t ur e are being treated

as separate entities, they are clearly closely related to each other. It might, in fact, be useful to conceive of hopes and fears as the positive and negative limits on the range of forms expectations for t h e

f u t u r e c a n a ssume. I n s o m e c ases, therefore,

expectations may coincide with hopes or fears. Even when this is not the case, the content of one +ye of image influences the content of

t h e o t h ers, an d al l t h r e e ar e a f f ected by b r o ader

structural conditions, as well as by other beliefs, values and ideas prevalent within the society. THE SELF-ANCHOiuNG STRIVI NG SCALE

The hopes and fears of the South African elites were isolated with the aid of the Self-Anchoring Striving Scale developed by Hadley Cantril (i96g) and Lloyd Free. The scale rests on two questions which ask th e i n t erviewee to d escribe the best an d w orst possible future h e c a n i m a g ine. T h e " b est f u t u r e" , o r

future hopes, forms the upper limit o f

t h e t en-point Self-

A nchoring Striving Scale, while the " w o rst f u t ure", o r f u t u r e fears, forms the lower l i m it . O nce th e end-points of th e scale

or "ladder" are defined in this manner, individuals are asked t o rate their c o untr y i n t e r m s o f i t s p r esent position o n t h e

scale, its position five years ago, and its anticipated position five years hence.~ Through the use of t h e Self-Anchoring Striving Scale it is possible to let respondents express their concerns with regard to t h eir country's future w i thout prestructur-

ing their rephes. At the same time the various scale or ladder ratings of diHerent groups can be compared, as can the content of expressed hopes and fears. In o r der t o f a c i lita.te the second

aye of comparison, Cantril has dehneated a broad spectrum of content analysis categories which serve to define the elements +In order to tie the Striving Scale questions more closely to other qnetions dealing with the future, respondents in the elite study were also asked to estimate their country's position on t h e l a dder t wenty years hence. Consequently, the original scale was somewhat expanded for the purposes of this study.

1MAGKS OF THE F LTLrRK

i53

of any individual's hopes and feats. The major categories used are as foHoww: Political, Social, Economic, International, Status of Nation and GeneraL A given respondent can articulate hopes or fears that fall into one or m ore of these groups. The eli te ' r e sponses to t h e q u estions on h o p es an d f e ars

which underlie the Self-Anchoring Striving Scale exhibit a number of genetal patterns. To b e gin w i th , t w o -thirds of t h e respondents articulate their h o pes for th e f u t ure i n s o cial terms,

while 35 to y5 per cent express political or economic hopes. Similarly, close to half of the respondents have fears of a pohtical nature, whereas fears concerning national status or social

problems are articulated by ao-yo per cent. In addition, very few respondents are concerned about developments in the inter-

national sphere. (See the 6rst row of tables z8 and r9.) If hopes a nd feam are viewed in more general terms, rather than in t h e

context of the specific content analysis categories, respondents are most likely to articulate their hopes in terms of i n ternal or

external peace and harmony (35-g5 per cent are in this group), whereas the vast majority are fearful that the f u t ure will b r i ng

disruption of present stability. (See right-hand column of table When the third part of the Self-Anchoring Striving Scale (the ladder rating) is related to hopes and fears, it becomes clear that the vast majority of respondents have a high level of expectations for th e f uture and ar e u n quali6ed optimists. This means that their future ladder ratings are closer to their hopes or at the

upper part of the ten-point Striving Scale. (The mean rating for 6ve yeat3 into the future is 6,5 and is y,5 for twenty years hence.) Furthermore, ratings tend to move steadily upwards as the tetnporal focus shifts. That is, the mean rating for South Africa's position 6ve years ago is approximately rung 5, the present rating tends to be rung 6, the country is usually seen at rung 6 or p in 6 v e years' time, and th e mean rating fo r a v enty yeat3 hence is close to rung 8. I n t h e l ight of t his, the dire nature of

the elites' apprehensions, dealing with unre t and disorder as they typicaHy do, is less starthng than it initially appears. For, since all segments of the elites have a uniformly high level of

a nd believe the future will be a f a i rl y accurate reexpect ations flection of their hopes, they do not set great store by their feats.

This considerably mitigates the signihcance of their negative character. In addition to the broad patterns of similarity outlined above,

there are, of course, difference among various elite groups. The most pronounced divergences are usuaHy in terms of the tra-

I54

IMAGES OF THK FUTURE

ditional South African polariser — home language. The two major groups can best. be differentiated as foHows: I. In terms of specific future concerns, Afrikaans-speakers are most likely t o e xpress hopes and fears dealing with t h eir country's national status, while English-speakeis are more likely to be concerned about social matters, especially in relation to their fears.

r. On a more general level„ the majority of Afrikaans-speake rs hope that th e f u ture will b r in g a

g r eater realisation of t h e

government's separate development policy, whereas they fear that existing stability will be upset by some sort of foreign intervention o r

p r e m a ture i n t egration, a s t a b l e 2 p i nd i c ates.

English-speakers, in contrast, are more hkely to hope for greater racial integration or increased opportunities for all groups to develop their potential, while they tend to articulate their fears in terms of an unspecified threat to existing stability or greater repression and discrimination by the government. 3. Afrikaans-speakers tend to have higher ladder ratings than then' English counterparts, which means their for

expectations

the future tend t o b e closer to t h eir h opes than to t h ei r f eai3.

The English-speakers, in contrast, are more likely to exhibit a l ow level of expectations, although the majority do n ot. I n addition, they are more likely to be pessimistic about the future

than are the Afrikaans-speakers, of whom the vast majority are unqualified optimists and foresee steady positive change for the future, as the range of mean ladder ratings illustrates. Afrikaners expect theircountry to have advanced an average of two r ungs on th e S t r i ving Scale i n t w t mity yeais' t i m e (the mean

present rating for this group is 6,2 and the rating for twenty yeaix hence is 8,r ). Englishon the other hand, expect their country t o

speaker,

a d v ance less; on t h e a v erage they f o resee it

moving from a present position of g,3 to a r ating of 6,8 in twenty years. These differences between the major lallguage groups are, however, relatively minor. Regardless of home language, the majority of respondents expect the future to bring considerable realisation of their hopes (i.e., their ladder ratings for the future are at rung 6 or h i gher

) and they foresee positive

change for South Africa. This uniformly high level of expectations among the eHtes is rather surprising in v iew o f

t h e f a c t t h a t sizeable segments of

the two major l anguage groups have diametrically oppoW hopm and fears, That is, those Afrikaneis who hope for separate development and fear premature integration are q u ite un-

like the English-speakers who want to see greater racial inte-

gration, but fear greater repremon and discrimination as a result of present government policy. Yet, the majority of respondents espousing either of these opposite patterns of hopes and fears have a high level of expectations. Most expect their country to have made considerable progress in achieving their hopes in twenty yeats' time, although the group hoping for separate development, but fearing integration, does have somewhat higher ladder ratings than the group displaying the opposite pattern of hopes and fears.+ It would appear, then, that the South African respondents tend t o v iew t h e l i k ely f u ture through the prism of their hopes, since two completely opposite visions of the future cannot be equaHy likely, given the constraints of the present. It should, of course, be noted that the contemporary situation in South Africa is an extremely complex one containing various conflicting trends. For, while integration proceeds apace in the economic sphere, separate development is becoming more of a reality pohtically and territorially. Consequently, the respondents' tendency to expect their hopes to be realised may not be wish-fulfilment so much as a process of assessing the long-range significance of present trends dBerentiaHy. That is, if one believes economic trends hold the key to the future, one expects racial integration to increase in the years to come. If, on the other hand, one sees political decisions as paramount in shaping the future, one expects greater separate development to materialise. Given the countervailing tendencies of the present, therefore, it may not be at all unusual that some elements of the South African elites see opposite images of the future as equally plausible. Attitudinal differences similar to those outlined above also hold for t w o v a riables most closely associated with h ome language: religious and political preference. That is, members of the Dutch Reformed Church and the National Party tend to have hopes, fears and expectation levels similar to those of the Afrikaans-speakers. English attitudes, in contrast, are found among respondents belonging to other religious and political groups. This indicates that the various eitxnents of the SelfAnchoring String Scale have similar causal antecedents. More detailed analysis showed that the general form of hopes and fears is directly affected by religious and political preference, +The group hoping for separate development, but fearing premature integration, expects South Africa to be around rung 8 in twenty years' time, whereas respondents hoping for rmial i ntegration and fearing governmental repression or discrimination envisage the country at rung 6 at the same point in time.

Ig6

IMAGES OF THE FtrrURE TA E I

E 27

anguo r

General JYarnroJ' o Hopor and Fears by &orna L Horne Language

Total

Both H OPES FOR TH E

FVTLRE

Internal/External Peace and Harmony Separate Development

Opportunities for AH Integration Total Number Qf Cases, . Foreign I ntervention Premature Integra-

38.0

45,4

48,2 12,4 1,4

32,0 15,5

7P

12,1 24,0 24,2

100,0

100,0

100,0

100,0

(363)

(97)

(396)

(856)

33,9

27,8

18,5

tion

General Threat to Existing Stability Repression anti Discrimination No Fears Total umber of Cases ..

39,6 29,7 18,1 12,6

14,7

20,0

20,9

227

24,5

22.7

4,6 14,9

11,3 18,6

25,7 16,5

15,1 16,1

100,0

100,0

100,0

(363)

(97)

(396)

100,0

(856)

Future Hopes: Cramer's U = 0,324 (pos.) Future Fears: Cramer's V = 0,228 (pos.)

while ethnic bacltground and home language have an indirect impact through their influence on the first two variables. EXPECTATIO.'sS FOR THE FUTURE

In any country certain aspects of the present pose greater obstacles for the future than others. Within the South African context four areas may be especially problematic during the next twenty years: the nation's governmental form; the degree of control over individual freedom exercised by the government; the nature of foreign relations, and the importance of r acial o r

c u l t ural d i f f erences in societal a f f a irs. Gn e w a y o f

assessing the possible importance of any of t hese matters for South Africa's future is to determine the extent to which elites differ in their attitudes towards than. If elite respondents are

lMAGES OF THE FUTLKE

t5'7

sharply differentiated on a given issue, it cm be considered problematic, while if most respondents share similar ideas about an aspect of the future, it wiU not be of great import. Consequently, elites were asked to

o u t l ine t heir f u t ur e e xpectations

for each of the potential areas of conflict mentioned above. Their responses indicate that the 6rst two areas may be problematic, whereas the latter two may not .

Analysis of the questions dealing with future expectations for the four f ocal areas outlined above indicates that the elites

are relatively sharply divided in terms of their ideas about parliamentary representation and g overnmental controL T hese differences can best be summarised in terms of t h e ethnic cleavages within the sample. Table a8 shows that the majority of A f ~ s-sp eakers expect various racial groups to h ave separate national parliaments, whereas the English-speakers are more likely to foresee a common parliament for all groups. The second group also expects the future to bring lessened governmental c ontrol o f

t h e i n d i v i dual's f reedom o f c h o i ce.

The majority of the Afrikaans group, in contrast, does Not expect lesser legal controL Differences of this type indicate that there is considerable disagreement among the elites about the type of parliamentary representation and governmental control the future will bring. Consequently, these two areas may indeed generate problems in the years to come. For the remaining two focal areas considered, the South African elites are quite united in their expectations, as table a8 illustrates. Irrespective of their various background characteristics, about two-thirds of the respondents expect improved foreign relations with the A f rican states, while they do not foresee much change in South Africa's p rin t r elations with the communist countries. This uniformity of opinion also prev ails in t h e r e spondents' statements about t h e f u t u r e o f t h e

country's inter-group relations. The majority expect lesser distinctions and closer relations between the major White groups, as well as between the racial groups. These trends serve to illustrate the fact that the elites are in broad agreement about

the type of foreign relations and cultural cleavages the future holds in store. This is not to say that these areas wiH not, indeed, give rise to future diSculties. In fact, they may prove to be the most problematic, precisely because the elites do not seem to believe that a variety of developmental paths are feasible. C outh Africa can be caught much more unawares in these two areas (Le., foreign relations and internal cultural differences) than it can in terms of parliamentary r~

ogently, S

Igs

IMAGES OF THE FUTURE TAEI.E

28

ExPcfkrlionsfor the Fatale by Hrrrar Language

Home Language Nanlre of Expectation

Total NATURE OP PARLIAIIENTARV REPRESENTATION

Percentage expecting comInon parliament

No. of cases on which per cent based DEGREE OP CA)~

36,0

51,9

Tau = 0,283

(p < 0,001)

(356)

T AL

CON fROI O P KVDIVIDUAL

Percentage expecting decreased control No. of cases on which per cent based

28,5

49,1

(351)

Tau = OP03

(p < 0,001)

(818)

F OREIGN R E L ATIONS W r f H

Percentage expecting positive change

77,6

77,9

($70)

(98)

(39$)

20,1

22,9

29,5

77,4

No. of cases on which per cent based F OREIGN RELATIONS Ivfr a CO ~

Tau = 0,011

(p < 0,10)

C OUIff R IES

Percentage expecting positive change

No. of cases on which per cent based

(364)

Tau = 0,09$

(p < 0,001) (387)

(847)

D ECREE O P E U ROPEAN I N TERACIION

Percentage

expecting

greatcf lntcfactlon

No. of cases on which per cent based INDEX O P TIQNS

R A C IAL DISPEL

Percentage cxpectmg lcsscf distinctions

No. of cases on which per cent based

70,0

Tau = 0,128

(p < 0,001) (98)

(402)

(869)

67,7

56,5

62,8

(96)

(395)

(855)

Tau = 0,119

(p < 0,001)

The statistic used to measure degree of association is Kendall's Tau

I MAGES OF THE FUTU R E

I59

presen tation and governmental control where a number of different future possibilities are current. In addition to analysing the nature of the elites' expectations in a n u m ber o f s pecific areas, expectations were also rel ated to at titudes on t h e same subjects. Th e v ast m ajorit y o f

respondents have similar attitudes and expectations (i.e., they consider their future expectations in any given area desirable). This is, however, somewhat more true for certain areas than for otheIs. The percentages considering developments desirable in the resp+:tive areas are as follows: Foreign rdations with Africa 96 Afrikaans-English relations 95 Interracial Ivlations 86 Foreign relations with communist countries 85 Parliamentary representation SR Governmental control 75 The congruence of attitudes and expectations is most complete

for foreign relations with Africa and interaction between the White ethnic groups, while attitudinal and expectational divergeno s are greatest for governmental control. The last variable is also the only one for which the totals for attitudes and expectationsdiffer to any appreciable extent. That is, the majority of respondents (57 per cent) favour decreased governmental control of the individual, but 6I per cent do not expect this to become a reality. In spite of these differences, 75 per cent of the sample have similar a t titudes and expectations concerning

governmental control. Because of the m arked similarity between the two t~~ o f v a r iables, it was generally impossible to determine which one causes the other. For those two variables (i.e. parliamentary representation and governmental control) where some semblance of a relationslup emerges, it appears t hat a t t i t u des l e ad t o

ex p e c tations r a t her t h a n v i c e

versa. By and large, however, both variables tend to be identical. I MAGES OF THE Ft:TURK AS CAUSAL V A R IAB L E S

Thus far this chapter has treated images of the future solely as dependent variables. We have seen that insofar as elites have differing ideas about the future, these attitudinal differences are conditioned by ethni, political and rehgious afBiations. That is, the four ethnically hnked background variables (e.g., ethnic background, home language, religious preference and political preference) tend to mould the images of the future articulated

t6o

IMAGES OF THE FUTURE

by the dites. Further data analysis, however, reveals that, once formulated, images of the future have a causal impact of their own. It appears that hopes and fears for the future help to define attitudes towards a number of present situations such as how

much segregation there ought to be. It wiH be recalled from the discussion in chapter 4 that respondents' scores on the segregation index are shaped by a number of background characteristics. More specifically, home language, religious preference and political preference have a direct impact on the level of scale

scores, whereas ethnic background has largely an indirect impact via its influence on the other background factors. Introduction of the future variables, however, reveals that hopes and fears for the future intervene in relationships between the background variables and the segregation index. Future hopes and the segregation index are fairly strongly telated. Of the respondents hoping for separate development, 67 per cent favour moderate or pronounced legal segregation, while only 3 per cent of those hoping for some form of integration have mnilar views on segregation. Respondents hoping for greater opportunities for aH groups or internal/external peace and harmony, fall in between these two extremes. Fears for the

future exhibit a pattern similar to that observed for hopes although the relationship is weaker. The majority of respond-

ents fearing foreign intervention and premature integration favour moderate or pronounced legal segregation (5z and 5g per cent respectively). In contrast, 65 per cent of those apprehensive about a general threat to the existing stability and 88 per cent of those fearing greater repression and discr e t i o n f avour limited segregation. The significance of these relationships between the segtegation index and hopes and fears emetges best

when the impact of the background characteristics is also con-

sidered. Controlling for hopes and fears tend to reduce the association between these variables and the segregation index, as the pattern of relationships for political preference iHustrates. When hopes for the future are held constant. the degree of association between political preference and segregation scores is reduced to a Mu e of o,4t4 on the average. Since the Tau value

of the original relationship is o,6t t, controlling hopes brings about a 3a per cent reduction.+ This implies that hopes intervene in the relationship, although political preference con+If one ignoresthe group hoping for greater peace and harmony, which is in some sense a re idual or "no answer" category, the reduction of the relationship is even greater. Yau takes on an average vaiue of original 0,342 in this case, thereby reducing the association between pohtical preference and the segregation index by about half.

I MAGES OF THE FUT U R E

I6I

tinues to have a meaningful impact on attitudes towards legal ~g at i on, as does home language. f H o l ding feaI3 for th e future constant also reduces the relationship between political

preference and attitudes towards segrqption. The drop is considerably less than for hopes, however, since the degree of association is only reduced by I8 per cent (Kendaii's Tau drops from o,6I I to o,5oz). Consequently, fears appear to intervene less in th e r elationship between background and a t titude variables than do f u t e h o pes. This more limited impact of fears is in keeping with the earlier finding that respondents' expectation levels are more strongly influenced by their hopes than by their fam (see pp. I5y-g). In s um, it a p p ears that h opes and f e ars for t h e f u t u re a re

moulded by the same ethnically associated background characteristics as attitudes towards segregation are. Hopes, however, and to a lesser extent fears, also have an impact on the nature of attitude scale scores as the foHowing diagram illustrates: Hopes for the Future

B ackgro und Characteristics

(and fears) Attitudes Towards Segregation

This pattern of rdationships implies that images of the future, once established, can also have an independent causal impact of their own. A COMPARATIVE PERSPECTIVE

Within the South African context the need for comparative analysis is crucial, since Inany elements of the country's internal

make-up and international position appear unique. Furthermore, the elites themselves occupy an unusual position within the larger society and, therefore, their attitudes may not be typical of the larger population. In v iew of t his, the intranational and international comparisons outlined below should serve to add perspective to the preceding discussion of the South African elites' images of the future. Ertfernnl corrtpurisorss

In Ig6g, William Hudson, Gideon Jacobs and Simon Biesheuvel undertook a lengthy investigation of

t h e basic attitudes

)ControOing for hopes and fears reduces the impact of hornelanguage

on attitudes towards segregation to the same extent that it weakens the

inftnence of political preference.

re

KWACES OP 'HIE FUTuRE

and values characteristic of the larger White South African population. The authors initiated their research because they believed that knowledge of present attitudes (or hopes and fears) would facilitate prediction of people's reactions to future events a nd policies (Hudson, Jacobs and Biesheuvel, i966: i z). T h e study consisted of two major parts: ( i ) a large-scale opinion poll dealing with political issues and utilising a representative

sample of i ooo interviewees, and (c) a series of in-depthinterviewswith a sample of i oo. The second part of the study, conducted between i96S and i965, is of primary interest in the present comparative context. Although the smaller sample was not chosen randomly, and is biased in favour of leadership-type pursuits, it includes a much wider range of respondents than is the case with

t h e e h t e s ample. T h e i n t e r views were r a t h er

lengthy and most questions were answered in considerable detail. One of them dealt with the future and asked about the shape of South Africa in ten, twenty and fifty years' time. In their general ~ men t of ideas about the future, Hudson, Jacobs and Biesheuvel conclude (i966: g8): "Nearly two-thirds express op~ or qua l ified optimism about the future. Of these a minority either paint a msy picture and appear to have no fears or doubts about everything coming out all right, or they adopt an idealistic view which approaches the brotherhood of man and sees ail races in South Africa living together in harmony to their mutual advantages. The qualified optimists temper their hopes by a realisation of the possibiTities of internal and external attacks and Thirty per cent are pessimistic. They foresee bloodshed, Sharpevilles and chaos." The above characterisation could also apply to the elites' attitudes in that g6 per cent of them are pessimistic about the future (i.e., their future ladder ratings are lour than those for t he p~ n t ) while the remaining 7@ per cent are optimistic (i.e. future ratings are higher than scores for the present). In terms of more specific images, Hudson, Jacobs and Biesheuvel ,'i966: 9 7-i' ) found that the majority of their respondents foresaw a continuation of present trends and practices for

aggres sion.

the next io years, That is, they expected the government to con-

tinue its efforts to implement separate development. Only about ag-3o per cent, believed these efforts would continue for as long as twenty year's, however. This sinaller group anticipated contin ued development of the Bantustans and withdrawal of t h e Africans from "White" t owns. I n a d d ition, they expected

MAGES OF TH E F U TURE

%hite South Africa to lose both land and a good part of its labour force, while retaining considerable political control. Such an image of the future is quite similar to that formulated by those elite interviewees hoping for some measure of separate

development. This latter group also composes about 3o per cent of the larger whole. In contrast to those anticipating some form of separate development, Hudson, Jacobs and Biesheuvel isolated a group foreseeing a greater degree of integration. These respondents believe that as concessions are made to Blacks in the economic and political arena, tensions will lessen and greater interaction will occur. Although the speed with which political integration is to proceed varies, most respondents believe it will develop slowly and begin to take hold within the next ten years. This group is quite similar to the i S per cent of the elites hoping for greater racial integration. In terms of fears for the future, both the Hudson-Jacobs-Biesheuvel sample and the elite sample are again similar. In both cases the same spectrum of fears are expressed. In the earlier study three types of fears are most prevalent. disaster and chaos arising from the institution of a Black government; racial strife due to failure of the government's present policies, and a demise of Western civilisation for a variety of reasons. These parallel the fears expressed by many of the elites. They fear that the presently existing stability might be threatened by premature integration (ao per cent). that there might be continued repression and discrimination on the part o f t h e government (i5 per cent), or that some unspecified threat might lead to chaos and anarchy (z3 per cent). It should be noted that a quarter of the elites fear some type of foreign intervention, while very few of Hudson, Jacobs and Biesheuvel's respondents do. In pa rt this may be du e t o th e fact t ha t m a n y f oreign countries (e.g.,

Communist China, Zambia, East Africa) had become more hostile in th e t h ree years between the tw o studim. C onsequently,

the elites may have been more aware of outside threats because they were questioned at a later date. It may also be that the elites are in positions that make them more aware of outside dangers; certainly they are less pmmncial than the average South African. The subject of foreign intervention is the only respect however, in which the hopes and fears of the two samples seem to diverge. In addition to inquiring about general images of the future, Hudson, Jacobs and Biesheuvel (i966: 33-g8) almi asked their respondents about future developments in a number of specific

I64

IMAGES OF THE FUTURE

a reas. In terms of political representation, over half of t h e sample expected separate voters' rolls for all groups, m~ that each would have its own representatives in the central parliament., although their number would vary. Of the retnaining respondents,about Ig per cent believed everyone would be on the same voteIs' roll, but a qualified franchise ought to prevail, and 29 per cent anticipated separate parliaments for all groups in keeping with the policy of separate development. The elites' expectations for parliamentary representation exhibit a

s i m i lar

pattern: 4,7 per cent expect a single parliament with separate rolls for one or more Black groups, 5 per cent, foresee a common voters' roll for al l g roups, and ~ p e r c ent anticipate separate parliaments for at least one Black group. Similar patterns also emerged in a I9)2 survey of a representative sample of the whole White population. The study was undertaken by Market Research (Pty.) I.td. for the newspaper, support, and found that the dominant emphasis in aH groups questioned was on separate voters' rolls and separate parliaments as the most likely forms of political representation in the future. A small number of respondents (ranging from I to Io per cent in different categories) foresaw representation through a common parliament with a single voters' roll (see Rapport, November I 2 aud I9~ I9)2 ).

In addition to parliamentary representation, Hudson, Jacobs and Biesheuvel questioned their interviewees about a number of o ther issues. In th e a rea o f

i n t ernational aff aiIs, much of t h e

sample was optimistic with regard to the future, which is in keeping with the elite data where over three-quarters of the respondents expect improved relations with the rest of Africa. A similar optimism tends to pervade ideas about relations between the two White groups. Seventy per cent of the Hudson-JacobsBiesheuvel sample believe that the A f r ikaans- and Englishs peaking groups have been moving steadily closer to o n e another and most favour an eventual merger of the two groups.

The elites have similar attitudeI; 7o per cent expect greater interaction between the White groups in the future. T he above discussion in dicates that t h e

g r o ups studied b y

Hudson, Jacobs and Biesheuvel are remarkably similar to the elite sample in terms of f u t ure expectations, as well as in terms

of hopes and fears. Of course, the above comparisons are, at best, scanty and can only serve to indicate possible associations. Nonetheless, it seems that the elite respondents have ideas about the future of South Africa that are largely the same as those prevalent within the larger White population. This would imply

XMAGKS OF THE FUTURE

that those compromises and pohcies which the elites might accept in the future would also meet with the approval of the average citizen. Internationalcorngarisons

During the late 6fties and early sixties the Self-Anchoring Striving Scale was administered to more-or-less representative samples in fourteen countries. The diveme assortment of countries studied included: Brazil, Cuba, the Dominican Republic, Egypt, India, Israel, Japan, Nigeria, Panama, the Philippines, the United States, West Germany and Yugoslavia. Because of inadequacies in some of the data only twelve of these count ries consistently lent

t h e mselves to c o m parisons w it h S o u t h

A frica.+ The data generated by t h e adxxtuxistration of t h e Striving Scale in these countries are presented in some detail in Hadley Cantril's book, The Pattern of Human Concerns(x96gI. When compared to other countries, the South African pattern of hopes reveals some interesting peculiarities, as table ug shows. It has the smallest proportion of respondents expressing economic concerns and the percentage of elites expressing international hopes is relatively low too. In contrast, the South African group ranks second in terms of degree of social concern and i n terms of t h e percentage of r espondents hoping fo r a

main-

tenance of the status quo. South Africa's level of political and national status hopes tends to be somewhat above average. In

suxn, the country is most hopeful of achieving social betterment, but is relatively unconcerned about economic and international relations. This is in keeping with the general South African view that the country's most important challenge lies in the arena of racial and ethnic relations. Continued economic advancement tends to be taken for granted, whereas international relations are deemed of little consequence for South Africa's future. It i s w o r th n o t i ng t h a t t h e S o uth A f r i c an p a t t ern o f

hopes is most similar to that of Nigeria, the only other African country in which the Striving Scale was administered. Table 3o indicates that fears for the national future tend to be more dispersed and diverse than hopes. When compared with other countrim, the South A f r i can pattern of f ears is a bit

unusual. The elite sample ranks 6mt in terms of the proportion of respondents expressing social fears and in t erms of those stating they have no fears for the future. It aim has the second h ighest proportion o f

r e s pondents expressing fears a bout t h e

For the Japanese sample the content of national hopes and fears was not available, but ladder ratings were.

166

IMAGES OF THE FUTURE

va s t . a 29 Por ant of JAtfonol SontpfosExpress ing Various HoPos Nature of Concern

Economic South A&ica (N = 888) Brazil

58

(N = 2739)

Inter- Statusof national Nation

Political

Status

Quo

19

Cuba

(N

=

1490)

Dominical Rep. (N = 2442 )

78

(N = 1237) India (N = 5720) Israel (N = 1170) Nigeria (N = 2876) Panasna (N = 1351) Phibppines . (N = 1388) United States (N = 2 696) I,'Vest Germany (N = 480) Yugoslavia .. (N = 1523) ,

70

69

50

*

13

52

13

59 42

81 70

Senror: I'able Vill t 26 in Cantril (1965: 180).

country's national status (only Egypt has a higher p«portion) In contrast, the South African level of fears concerning international affaixs is lower than that of any other country and the elite is second lowest in terms of economic fels. The level of pohtical fears fends to be somewhat high in relation to the other countries. It will be noted that the pattern of feats is quite ~ to tha t p r eviously outlined for national hopes. ~~ on6rms the earlier conclusion that South Africans view

arena, with its complicated racial picture, with greater concern than do other countries. The pattern of fears also shows that economic advancement tends to be taken for granted and the international alanna is considered unimportant

The high degree of apprehension displayed with regard to South Africa's national status is undoubtedly a reAection of outside threats to solve the country's problems through force-

BGlGES OF THE FUTURE

ful intervention. It is important to note that the pattern of fears

most closely related to South Africa's is again that found within the Nigerian sample. Both of these countries are located within Africa and have quite diverse ethnic and racial compositions. These traits may prrEent them with similar future prospects which, in turn, lead to similar patterns of hopes and fears for the future. vasss 30 per Centof polionnl Sunipks E xpressingVmions Pears Nature of Concern International Political

South Africa (N = 888) Brazil (N = 2 739) Cuba (N = 1490) Dominican Rep. (K = 2442)

Econonuc

Status of Nation

50

59

egypt

(N = 1237 ) India (N = 5720) Isrml (N = I 170) Nigeria (V = 2 876) Panama (N = I 351) Philippines .. (N = 1 388) United States (N = 2696) 4'est G~y (N = 480) Yugoslavia ..

20

17

23 19

(N 1 523) Average

19 sonrsrs Table Vlii t 27 in Cantril (1965: 181).

Table 31 presents the mean past, present and future ladder ratings for each of the countries in which the Striving Scale was administered. South Africa, it w ill b e noted, has one of the lowest mean ratings for the future (only Japan and the Philippines have lower ratings), while its past rating is somewhat above average and its present rating is very close to the average. Although it is difficult to assess the meaning of relatively high or low mean ratings, the shifts in scale scores from past to pre

t68

B CAleES OF THE F ~

sent and present to future are amenable to fairly clear interpretation. The right-hand columns of table gt enumerate the shifts in m ean r atings for each country. T hey show t hat m ost countries endamage a shift of t to a r u n gs as t hey m o ve f rom past to present or from present to future (i.e., the mean ratings

differ one to two points). South Africa does not conform to this trend in t hat its eiites envisage a somewhat below average degree of past p.ogress and anticipate a ver low level of future advance. Only the United States is as low as South Africa in terms of the shift from present to future. This relatively conservative evaluation of past and, especially, future change could be due to the fact that South Africa is relatively advanced industrially and has a government that appears to be having second thoughts about continued rapid economic development.+ It could also be that South A f rican conservatism stems from greater awareness of the nation's problems than is typical of many of the countries in which the Stri~qng Scale was administered.

Although the preceding discussion implies that South African images of the future are relatively unique, this is diRicult to assess without some conception of how the p a ttern of f u t ure

concerns evolves as a country becomes more economically developed. An effort was made to arrive at such a conception by relating three social indicators to the pattern of future concerns and the ladder ratings. The analysis revealed that as a country's level of living, literacy rate and degree of urbanisation increase, so does the breadth of its future concern. This progressive change makes itself visible in two ways: the scope of concerns increases (that is, a majority of respondents express concerns in more areas), and the level of economic concern decreases, while the level of concern in other areas increases. A lthough t h e s o cial i n d i c ators ar e p o sitively r e l ated t o t h e

breadth of future concern, they are negatively related to the degree of future advance anticipated (as measured by the ladder ratings). As a nation's level of living, literacy rate and degree of urbanisation increase, the degree of future progress anticipated by its citizens tends to decrease. It would appear that as a country becomes more educated, urbanised and weil-tMo, its citizens develop a broader concern for the future and become aware of a wider range of developmental pombilities. Concommitantly, the complexity of change is perceived more clearly, +These second thoughts crop up periodicaify when government spokesmen note that realisation of separate devefopment is more important than ~n o mic development and may require temporary delays in industrialisation.

169

IMAGES OF THE FUTURE s Ant.z 3 1

Larbiar tati>rgs by Gentry r Mens and Shigis Shirts in Ratings

Mean Ratings

PresentFuture

Future

South A&ica (N = 888) .. Brazil (N = 2 739) Cuba (N = 1490) Dominican Republic (X = 2 442) Egypt (N = 1 237) India (N = 5 720) Israel (N = 1 170) Japan (N = 972) .. Nigeria (N = 2876) Panama (N = 1351) Philippines (N = I 388) United States (N = 2696) West Germany (N = 480), . Yugoslavia (N = 1 523)

6,5

3,5

5>1

7,6

7,0

8,8

27

7,0

1,6

5,5

4,0

2,5

1,0

5,9

3,5 4,0

0,2

l>4

1,8

1,5

2,0

8,2

2,0

5,0

6,0

1>0

6,1

5,1

— 1,0

6,5

1,0

0,2 2>1

6,8

8,6 1,4

1,8

SaarrarT ables IX: 2, 3, 4, 5 in Cantril (1965: 18~7 ) .

leading to anticipation of more limited future advance. (For a more detailed discussion of the analysis leading to these conclusions, see Huber» 1973> ch. /. )

In terms of the considerations just outhned, South African elites do not deviate from the dominant trend. It appears, therefore, that although South African images of the future difFer from those in a good many other countries, this is not unusual in terms of the country's literacy level, degree of urbanisation and level of living. This implies that the images of the future articulated by South African elites are much like those of countries with similar developmental levels and social contexts, In

view of this, it is quite possible that background variables similar to those influencing the genesis of South African images of the future also affect such ideas in other areas of the world.

CHAPTER X

WHITHER SHITS SOUTH AMECAP R ELATIONS BKM V K K N

L A N G U A G E GR O U P S

This survey of the South African White ehtes has given empirical evidence of the sociological structure of the elites. But perhaps more important than that, it has given insights into some of the social processes occurring among them and possibly in the White society which they lead. The elites do not form a homogeneous and unified group, the principal cleavage being one based on religion and home language. In the past the Afrikaans- and English-speakers differed not only in their religious a8iliation but also markedly in their socio-economic status. For although Afrikaners won pohtical dominance in Ig~y and have never completely aMicated political power since then, they were an economically under-privileged group with a rural rather than an urban base. They tended to be geographically concentrated, occupationally

undifferentiated and to earn lower incomes than their Englishspeaking compatriots. Common religious allegiance, an early predeliction for their own brand of Christian national education, and their history of sufkring at t h e h ands of Br itish imperiali~ built up a fertile soil for a common ideology and a sense of group identity and cohesion. The political, racial and e thnic attitudes characteristic of A f r i kaans-speakers in t h i s survey bear vivid testimony of this cohesion and unanimity. The English-speaking elites, on the other hand, had ethnically diverse origins, a wide spread of religious allegiance and have grown up Inainly from an urban base. Historically Englishspeakers have held the key to economic power because of the skills which they conunanded and the ability to contribute capital. Their education was separate from t h a t o f A f r i k aners and, whilst not in fused with any p ar ticular nationalist ethos, it d id set them ap ar t f r o m t h eir W h i t e f e llow count+m en . B u t

English-speakers show little gmup cohesion and have not produced a coherent ideology. The significance of this group in the present state of flux lies i n the close and rather more recent links which bind i t t o

the

main stream of Western culture. These links have been clearly I70

WH IT HE R W H ITE SOUTH AFRICA P

revealed especia0y in the English churches and universities. In the past leaders in these sectors have actually been expatriates. Even today when South African born leaders are beginning to emerge, the churches have institutional l i nks with

t h e i nterna-

tional community, e.g. the World Council of Churches and the world-wide Catholic Anglican Communions. English-speaking academics likewise are members of an international community, very conscious of their standing within it , an d consequently open to new ideas. In the commercial and industrial fields the sharp edge of international competition underlines the need for rationalisation.

The survey reveals most strikingly ho w t h e s egmented agencies of primary socialisation mould t h e t w o l a nguage groups, perpetuating differences. It i l l ustrates, too, that t h e initial educational and religious separation is often perpetuated in occupational sectors where one language group tends to predominate. It is possible in South Africa to grow up an Afrikaans-speaker and never from the cradle to th e grave to come

into effective contact with Knghsh-speakels. The reverse is also true. The segmentation of the agencies of socialisation and of occupational spheres insulates both groups from effective contact and acculturatio~ Th e ethnic differences ale channeHed into political life via Qxe two main political parties, which are ethnically based. The analysis of political attitudes revealed some significant differences in political culture between the two groups. The primary concern of Afrikaans-speakers appears to be with their own group. Freedom is seen as important for the in-group; it must always be bounded by limits set by the Calvinist ethic as interpreted traditionally by the group. Dissent is legitimate only when it does not threaten the integrity and material interest of the group. A political party is seen as a vehicle for the realisation of group life in many spheres. State enterprise is seen as advantageous. English-speakea, on the other hand, seem to be more concerned with the individual and his role in society. They react against government regulation and control and against state enterprise. They have a greater tolerance of dissent and are more concerned with consent than with will in their vision of denlocfacv.

Hopes and fears for the future differ signi5cantly too. Afrikaans-speakers ale more hkely to express hopes and fears concerning their country's national status, while English-speakers are more likely to b e concerned about social matters. The language groups disagree about the type of parliamentary re-

WH1THER 1VHXTE SOUTH AFRICA?

presentation which the future will bring and about the amount of government intervention and control which is desirable. The nature of these hopes, fears and expectations would appear to flow directly from the differences in the ideological commitment

of the two groups. The survey gave empirical evidence of diHerences between English- and Afrikaans-speakers in their estimation of the most admirable South A f r i cans. O nce more th e e t h nocentrism and u nanimity o f A f r i k aners stand i n c l ear c o n t rast t o t h e m o r e

diverse choices of the English-speakers. And choices tended to reflect the ideological tendencies of the two groups, with Afrikaners admiring those who stood for language group and racial separatism, and English-speakers admiring integrationists. In their estimation of the most powerful South Africans, both groups tended to exaggerate the power of the rival group. Thus Afrikaners most often named economic leaders and Englishspeakers named political leaders. From this one may perhaps infer a d e gree of r esentment. Th e E n g lish r esentment is al so

clearly reflected in their adverse opinion of the Sroederbond. That English economic dominance has acted as a goad and a spur is dear from the history of Afrikaner nationalism during the past half centruy. That A f rikaner political dominance is resented and found frustrating may be inferred from the relative

political alienation of the English-speaking group. To economic dominance the answer is proving to be education and endeavour in a relatively open and competitive field, backed by the power of the state. In an ethnicaHy structured society poHtical dominance isa refl ection of numerical dominance. There cm be no answer for the English-speakers, short of a change in the nature of the Afrikaner group. The most important finding of the survey in this respect is that p r eferred social t y pes ar e e t hnically a n d

r a c i ally d eter-

mined. Only the numerically smaH and politicaHy powerless English group gave any promise at aH of a non-ethnic, nonracial approach. The emergence of a broad South Africanism based on the nation state rather than on the primary group would appear to be a rather remote contingency at the moment,

especially in the Afril.aans group, where ethnocentric images of power and ~ bilit y are reinforced by a relatively high level of intolerance, the existence of w h ich i s c l early r evealed both

with regard to political dissent and with regard to poEtical interlnalTlage.

It is apparent from our d a ta, however, that this original language cleavage, once consistent for education, religion, place

I73

WHITHER WH tTK SOUTH AFRICA P

of residence, occupation, soci~conomic status, political aailiation and attitudes, is yielding to a more complex design. Afrikaners have modernised, becoming urban dwellers like their English-speaking fellow c o untrymen. T h ei r s o cio-economic status has risen and their occupational distribution has diversified. Church and school remain as islands of inviolate "Afrikanerdom"; the bureaucracy is predominantly Afrikaans-speaking; and ideological unanimity bears witness to past group cohesion. But there is interesting evidence of change. In the first instance, the clam factor has obtruded and there is evidence of the alienation of Afrikaans workers from Afrikaans elites. This survey also gives empirical evidence of some degree of anglicisation of Afrikaans-speakers among the elites. In this shift they tend not only to abandon their unilingual status, but to shift attitudes and values toward those of their English-speaking compatriots. Even more interesting is the inHuence of occupational sector on attitudes toward segregation. In both the professional and mass media sectors attitudes were not t ie d t o

h o m e l a nguage

in the characteristic fashion. Travel abroad would also appear to inHuence attitudes toward segregation. It is possible to conjecture then that changes in attitude can follow exposure to cross-cultural contacts, and to the urban industrial environment. Other ~ ch ha s shown that such shifts in attitude, for instance with regard to the family, are not con6ned to the urban areas.From the empirical evidence we have, we can do no more than conjecture that a shift of attitude is possible, and that it has in fact taken place under certain conditions.

The inHuence of urbanisation, occupational differentiation and mobility, education and secularisation are very recent. They have not yetbroken the group cohesion of Afrikanerdom or interfered in any very significant way with the inner cohesion, a like-mindedness born of shared tradition, environment and goals. It is this cohesive, almost familial group, which operates the centralized executive organs of state, and w h ich d ominates

a parliament whose proudest claim is that of it s sovereignty over every person and group in South Africa. "Die wil van die Yolk is die wil van God" is a doctrine which leaves little room for government by consent, especially as the concept of "volk" remains restricted to the primordial group. It is based on the idea of the right of the Afrikaner group to assert their will over all who dissent. In f a c t t h e electoral dominance of A f r i kaners

at a national level virtually excludes English-speakers from the decision-making process within the central political structures. The nature of the A f rikaner group tends to accentuate the

WH I T HER W H ITE SOUTH AFRICA.

efficacy of political and executive dominance, and of bureaucratic centralisation. And this dominance results in a certain who degree of political alienation among the Englisht ended not only t o s upport political parties other than t h e National Pmty in power, but also to be much less politically active than Afrikaans-speakers. Their comparative apathy may well be the result of their political powerlessnesI. Recent research by Lever (Igp i) on the I97o election suggests that the split in the cra ne r r a nks revealed in the emergence of the H.N.P. resulted in a resurgence of political interest and activity on the part of English-speakeis. It is possible that once the processes of social change have eHectively brokenthe group solidarity of the Afrikaners, the whole White society wiB be freed for new political alignments, with the English-speaking group able to play a much more eHective and creative role than heretofore. New political alignments would give English-speakers the chance to bring more universal influences to bear on the closed parochial South African body politic. They could turn out to be the cultural brokers that help to introduce a new political dispensation. There is some evidence then not only of the adaptability of social structures (in the blumng of language groups and the inc~ in s t ructural diHerentiation ) but also of adaptation of attitudes which could presage much more far reaching change.

speaker,

ELITES AND THE MASSES

The socio-economic profile of the elite group confirmed that they diHered in many respects from the White population as a whole. They were predominantly male, the age structure bore no resemblance to that of the White population, and they were at the apex of the income and educational pyTamids. Both main language groups were represented in elite circles, but were unequally distributed in the various sectois, a factor which we have shown to contribute to the peculiar dynamic of South African politics. Religion is a factor of ethnicity, so all denominationsare represented, but their distribution, too, is skewed in various sectors. Owing to the paucity of research in South Africa, there is little comparable material which enables us to assess adequately the relationship o f

e h tes to ma sses with r e gard to a t t itudes.

In chapter 9, however, we record a remarkable concurrencebetween the findings of Hudson, Jacobs and Keheuvel (I966) and oursebes with regard to attitudes toward the future. In a recent pubhc opinion poli published in support (November is,

WHITE D

W H l r E SOUTH AFRICAP

tg7a) there is a similar concurrence. In the latter poll the language group differential is shown to be more important than that between rich and poor. In chapter g i t

wa s p o i n ted o ut t h a t a p p r o ximately o n e -

quarter of the elites feared foreign intervention, whereas very few had done so in the Hudson, Jacobs and Biesheuvel survey. But the time i nterval between the tw o surveys, an interval re-

markable for the swift development of international pressures on South Africa, makes it impossible to infer with certainty that elites are in fact less parochial and more conscious of the outside world than the population as a whole. In one area, that of political preference, the survey showed a considerable gap between elites and masses. Thus the Progressive Party, with a very small membership and little voting support, claimed the allegiance of rg per cent of the ehtes. A public opinion poll carried out in November tg6g showed that only p per cent of the voters supported this party just before the >g7o election. (Dagbreek, December si, >g6g ). Chapter g contains a demonstrable correlation between party preferred and attitudes toward segregation. One may infer that the more open racial attitudes and those more favourable to some degree of integration, are to be found more frequently at elite level than among the masses. B EYOND THE ELIT E S

There are two diametricall opposing views of the possibility of social change in South Africa. There are those who can only see violence, bloodshed and cataclysmic social revolution as the end of the continuing proem o f r acial polarisation and domination. In contrast there are those who see South Africa passing through a process of modernisation which will result ultimately in an affluent society in which the national income will be more equally shared between individuals and groups. (Houghton, tgyo). In a middle, and somewhat ambivalent position, is Heribert Adam (tgyt) who in Mo dernising Reciu/ Domination asserts that revolutionary change is unhkely to occur in the near future. He does not see this change coming about as the result of the demands of a rational economic system, and its potential for ameliorating the conditions of the subjugated masses. He postu-

lates the continued ability of the White minority to resist basic change because of"an increasingly streambned and expanding

system of sophisticated dominance". Given the racial and eth-

t)6

nic composition of the bureaucracy and the army he maintains that "a flexible oligarchy that calculates its chances of survival pragmatically and uses its means accordingly can cope without difficulty with internal opposition". (>gyt : t u 3). If change is to be expected he feels it can come only from the "graduaHy changing interest constellations within South Africa, or f r om war from without". (t97 t: r 3y.,' Limited changes only, in other words, can be expected to Row from the ffexibility and pragmatism of the r u l in g g r oup and f r om ,

a m on g o ther things, their

position of strength within the society. ( 19/I - 1 8 t ., M'hat light did our study throw on this debate P Our fmdings would certainly support Adam's thesis with* regard to the adaptabHity of the elite group. There is empiri cal evidence of some degree nf "anglicisation" of Afrikaans-speakers which need not necessarily be regarded as merely the result of culture contact. Afrikaans elites had changed their attitudes in response to the demands made upon them in certain sectors of society. As a group they are being subjected to the change in values which is characteristic of a modernising society. In the elite' p erceptions of the future there would appear to be room for further adaptations. AH elites accept the growth of commonly shared characteristics in the various ethnic and racial groups as being probable; most saw this development as being desirable. They feel the need for solidarity with Black Africa; they are in fact admitting that they are Africans, and that they have interests in common w it h p etsons of a d i f f erent

colour group. Klites are divided by language group on two issues in their perceptions of the future, namely on the constitutional forms of the future, and on the degree of government control which can be expected, and which is desirable. Such areas of dissent can be healthy, for they ensure a livelv debate which will increase the options open to the ruling group in whatever adaptations it has to make.

Furthermore, ehtes were almost unanimous in the Utopian images in stressing the desirability of racial harmony. Beyond t he schisms is common g r ound. T h a t , t o gether wit h c o m m o n

economic intents x~~H ensure the solidarity of the present ruling group. Even more interesting in the projections of the future was the

finding that

imagesof the future appear to be causal variables

i n attitudes toward r a cial separation. I f

t h i s w ere so it w o u l d

indeed constitute empirical support for Adam's observation with regard to the pragmatic nature of the ruling group. The peculiar

VAB T H XR W H ITE SOUTH AFRICA?

177

and strategic position of the English-speakers has already been pointed out. Thus the cleavage within the ruling group could be creative for the group as a whole ensuring a non-ideological pragmatic area of debate in which one group was able through its culture, economic and religious ties, to mediate new va1ues to the embattled White society.

There is evidence among White leaders of a growing awareness of the need for reinterpretation of the basic principles for o rdering race relations in S outh A f r ica. A n e w l o o k a t traditional ideologies was enhanced by an upsurge of critical g and self-criticism. During recent years increasing numbers of Af rikaans intellectuals, journalists, writers and even politicians have publicly criticised petty apartheid, colour discrimination, the lack of m eaningful contact between ethnic groups in South Africa, the lack of development in the Bantu homelands and the lack of scope for and low wage levels of Black workers.(See Hendrik W. van der Merwe, 197c: I n-y ). While conservatives, Nationalists and A f r i kaners have become

more sensitive to the injustices of colour discrixnination and petty apartheid in the common area, liberals and Enghsh-speakeIs have become more wilUng to consider the merits of separate development or grand apartheid, the development of the homelands, and the recognition of c ultural identity. It is t o d a y generally agreed by both conservative and liberal leaders that discrimination on the basis of colour and race must be eliminated while ethnicity or cultural identity is recognised. This is in conformity with current thought in the Western world. The major question that concerns us here is whether Whites are prepared to implement in full the principles of racial equality and fuH freedom of separate Black states as proclaimed by Nationalist leaders. The full implications of a genuine policy of ethnic d i f f e rentiation as d istinct f r om c o lour di s crimination

should not be overlooked, It should be noted that the ideal of separate states has provided for the escape from racial domination of only a minor proportion of the Black population. According to the most optimistic projections of conservative demographers and politicians, the majority of the Black population will remain de facto in the common society. The new approach whereby Bantustans are entitled to full nationhood and the right not to be discriminated against merely on the basis of colour, has two i m portant implications for those Africans permanently domiciled in the common area In the first place they should be entitled to full citizenship in the country of their domicile and, secondly, they should not be dis-

i78

WH ITHER W H ITE SOUTH AFRICA P

crimina.ted against on the basis of colour. The elimination of colour discrimination would thus lead to the establishment of a non-racial policy in the common society. We have presented evidence of a certain adaptability among the Whites and of an acceptance of social change as not only inevitable but also desirable. They accept the fact that nce and ethnic groups are moving closer to each other. But while there is evidence that ncw trends in at titudes have developed among

White leaders in recent decades, we have no reason to expect that fundamental structural change will follow solely from these

attitudes. Future behaviour cannot be predicted on the basis of knowledge of attitudes because behaviour is more often determined by social conditions and pressures of the environment than by attitudes. Regardless of people's attitudes, they tend to conform to the role-dictates of the groups to which they belong. Attitudes expressed by ehtes in this study are not accurate predictors of their future behaviour. The attitudes and opinions of the elites, as compared to those of the n i ~

, h ow e v er , are of

relatively greater importance in analysing trends in South Africa because of their position of leadership. Elites are leading the herd instead of running with it. Knowledge about their attitudes is thus important in asseming political trends in South Africa and in making projections for the future as political change may either be significantly inhibited or facilitated by prevailing attitudes. Attitudes play a major role in determining the facihty with which people adapt to changes which areforced on them. Heribert Adam (i97i) has demonstrated the pragmatic adaptations made by the White group in response to the demands of the new situation. But n one o f t hese adaptations have amounted to fundamental structural change regarding pohtical rights of Blacks in the common area. Economic and political self-interest, race and class prejudice among Whites block the program of Blacks toward full development and self-realisation in a social sense. (see Van der Merive, ig72: i i y - 8.) In spite of favourable attitudes expressed by elites and other Whitm, barriers to the social, political and economic development of Blacks remain. The removal of these barriers is dependent almost wholly on the bargaiziing power of the subordinate groups and need not be cic ely related to attitudes held by those in power. Internal pressures, sometimes in alliance with pressures from abroad, w il l

l a r gely d etermine th e n a t ure and

extent of fundamental di n g e i n South Africa.

BIBLIOGRAPHY Adam, Heribert tgjt M o d e rnizing Racial Domination: The Dynamics of South African Politics. Berkeley: University of California Press. Adam, Heribert (ed.) tg7 i S o u th Africa: Sociological Perspectives. London: Oxford University Press. Ashley, M. J. and H. W. van der Merwe 1969 "Academic Contrasts in South Africa." Sociology of Education 4a (Summer): a8y-agt. Reprint No. r of the Abe Bailey Institute of Interracial Studies, Cape Town.

Ashley, M. J. tg7o

"Education and Attitudes in South Af r i ca." Journal of

Education s (July): ag-3g "The Education of White Klites in South Africa." Comparative Education 7 (August): ga-yg. Reprint No. 6 of the Abe Bailey Institute of I n t erracial Studies, Cape Town. Bachrach, P. rg67 T h e Theory of Democratic Elitism. Boston: Little, Brown R Co. Baltzeil, K. Digby tg5g "'Who's Who in A m erica' and 'The Social Register': Elite and Upper Class Indexes in Metropolitan America." Pp. r7a-t8g in Reinhard Bendix and S. M. Li p set (editors). Glencoe: The Free Press. Bendix, Richard and Martin Lipset tgg8 C l ass, Status and Power. Glencoe, Illinois: Th e F r ee Press. Beth, Marian rgb "The Elite and the Elites." American Journal of Sociorg7t

Seym our

logy y7 (March): 7@6-75g. Blondel, J.

tg68 V o ters, Parties and Leaders: The Social Fabric of British Politics. Middlesex: Penguin. Bottomore, T. B. tg6g E l ites and Society. London: C. A. Watts.

Suitendag, J. J. and H. W. van der Merwe, rg7a "The Movement of Afrikaners into Higher Occupational Levels." Humanitas t: s97-go3.

Calpin, G. H. (ed.) tg58

The South African Way of Life. New York: C olumbia University Press.

Calpin, G. H. rg68 A t L ast We Have Got Our Country Back. Cape Town: Suren Publishers.

r79

Brrrr.roGRApav

r8o Cantril. Hadley

rg6~ The Pattern of Human Concerns. New Bruruwick, N.J.:

Rutgers University Press. Carter, Gwendolen M.

rgb

T h e Politics of Inequality: South Africa Since rg +. Lon-

don: Thames and Hudson. Carter, Gwendolen M., Thomas Karis and Newell M , S t ultz ro6) S o ut h A f r ica's Transkei: Th e Politics of Domestic Colonialism. Kvanston: N o r t hwestern University Press. Charton, X. C. J., M. J. Ashley and H. W . van der M erwe rg7o "Local Government Elite in South Africa." Pp. a3g-a 4l i n H. L . V 4 t t s (ed.) Focus on Cities (Proceedings of a Conference held in July r g68). Durban : U n i v ersity of Natal. CiBiers, S. P. rg7r A p p eal t o R e ason. Stellenbosch: U n i versity Publishers and Booksellers. Cronje, G. and J. D. Venter

rgb

D i e Patriargale Familie. Cape Town: HAUM.

Dahrendorf, Ralf rgb Cl a ss and Class Conflict in I n d ustrial Society, London : Routledge and Kegan Paul. Danziger, K. rg63 "Ideology and Utopia in South Africa : A M e t h odological Contribution to th e Sociology of K n o wledge." British Journal of Sociology r y: gg-76. Dexter, Lewis Anthony r gpo E l i t e a n d S p ecialized I n t erviewing. Evanston: N o r t h western Press. Du Toit, Brian M. rgb B eperkte Lidmaatskap. Cape Town: John Malherbe. Etzioni, A.

University

rg5q "The Functional Differentiation of Klites in the Kibbutz." American Journal of Sociology 6g: y76-4lp.

Gerstl, Joel and Robert Perrucci rg6g "Educational Ghannels and Elite Ivlobility : A C o m parative Analysis." Sociology of Education 38 (Spring); a24Q39.

Geyser, A. S., B. J. M arais, Hugo du Plessis, B. B. Keet, A. v an Selms and Others rg6o D e l ayed Action. Pretoria: Craft Press. Gey mn Pittius, K. F. W. rg3) Plaaslike en Sentrale Bestuur in Suid-Afrika in Teorie en P raktyk. Kaapstad: HA U M . rgyr S t a atsopvattings van die Voortrekkers en die Boere. Preteria: J. L. van Schaik. Guttsman, 4'V. L. rg63 T h e B r i t ish Pohtical E l i tes. London : M a c Gibbon and Kee.

x8x Halsey, A. H. xg6r "University Expansion and the Collegiate Ideal." UniversitiesQuarterly r 6: 5~-g8. Hammond, P. K. (ed.) rg64 Sociologists at Work. New York: Basic Books, Hanna, William John rg>o "Methodology, Technology, and th e S t udy o f A f r i can Elites," African Studies Review rS (April): gg-roy. Harris, P. B. rg68 I n t erest Groups in South African Politics. Salisbury: University College of Rhodesia. Hess R. D. and J. V. Torney r g67 Th e D e velopment of P olitical Attitudesin C h i l d ren. Chicago: Aldine. Higgins, Edward rg7o "The Sociology of Religion in South A f r ica." Paper dehvered at VI I W o r l d Congress of Sociology, Yama, Bulgai'la,

Hodge, Gillian M. M.

rgb

S o uth African Politics rq3S-rg3g: A Select Bibliography.

Cape To~m: University, School of Librarianship. Houghton, D. Hobart rg6y T h e South African Economy. Cape Town: Oxford University Press. rgvo "Y'hite Prosperity and white Supremacy in South Africa Today: A C omment." African Aff'airs 6g (October). Huber, Bettina J. ig73 Images of the Future Among the White South African Elite. Unpublished Ph.D. dissertation, Yale University. Hudson, William. Gideon FrancoisJacobs and Simon Biesheuvel rg66 A n atomy of South Africa: A Scientific Study of Present Day Attitudes. Cape Town: Purneii 8r Sons S.A. Hyman, Herbert H, rgb "The Value Systems of Different Classes: A Social Psychological Contribution to the Analysis of Stratification." Pp. @a6-~a in Reinhard Bendix and S. M. L i p set (editors',, Class, Status and Power. Glencoe: The Free Press. Janowitz, Morris rgb " S o cial Strati6cation arid the Comparative Analysis of Elites." Explorations in Entrepreneurial History 8 (Supplement): 6- r i. Josephson, Eric xg-,o "Resistance to Community Surveys." Social Problems i8

(Summer): i r7-rag.

Keller, Suzanne rg63 Beyond the Ruling Class. New York: Random House. Kuper, Hilda rg6o I n dian People in Natal, Durban: Natal University Press.

t8u

BIBLIOGRAPHY

Lasswell, H. D., D. Lerner and C. K. Rothwell rg52 T h e Comparative Study of Klites: An Introduction and Bibliography. Stanford: Stanford Univexzity Press. Lasswell, H. D. and D. Lerner

rgb

W o rld Revolutionary Klites. Cambridge, Mass.: M.I.T,

Press. Lerner, Daniel and Morten Gorden sg6g Euratlantica: C hanging Perspectives of the E uropean Klites. Cambridge, Mass.: M.I,T. Press. Lever, Henry rg68a "Ethnic Preferences of Immigrants." Journal for Social

Research tp (Dec.): s-tg tg68b Ethnic Attitudes of Johannesburg Youth. Johannesburg: Witwatersrand University Press. rg7t " The General Election : t gvo." Unpubhshed paper presented to South African Institute of Race Relations, July Ig7s. Lever, H. and 0, J. M. Wagner sg65 "Father's Education as a F a ctor A Hecting Social Di stance." Journal for Social Research (SA.) ty (June): t. sg6g "Urbanization and the Afrikaner." Race st (October):

st-t88.

Lipset, Seymour Martin and Mildred Schwartz tg66 T h e P olitics of Professionals. In Howard V ollmer and Donald L. M i l l s ( eds.) Professionalization. Eaglewood Cliffs: Prentim-Hall. Lipset, Seymour Martin and Aldo Solari (eds.) tg6y E l i t es in Latin A m erica. New York: O x f ord University Press.

Loubser, Jan J. sg68 " Calvinism, Equality and I n clusion: T h e C ase of Af r i kaner Cah~nism." Pp. 367-kg in S.¹ E i senstadt (ed.) T he Protestant Ethic an d M o d ernization. Ne w Y o r k : Basic Books. Louw, M. H, H. tg6g "'n Tipologie van Klites in Moderne Regering." Pretoria; UN ISA Inaugural Lecture. MacCrone, I. D. tgg7 Ra c eAttitudes in South Africa. London: Ox f ord University Press.

Marvick, D. (ed.) sg6t P o litical Decision-Makers. Glencoe: Free Press. Matthews, Donald R. tgg4 T h e Social Background of Political Decision-Makers. New

York: Doubleday.

sg6o U .S. Senators and Their World. Chapel Hill, N.C.: University of North Carolina Press.

BtttL1OGRhPHY

tg68

t8S

"United States Senators: A Collective Portrait." In S. C. Patterson (ed.) American Legislative Behaviour. Princt. ton: Van Kostrand.

Mills, C. %right t957 T h e Power Elite. New York : O x f ord U n i versity Press. Morris, James O. tg66 E l i tes, Intellectuals and Consensus: A S t udy of the Social Question and t h e I n d ustrial R elations System in Chile. Ithaca, N.Y. : N e w Y ork State School of I n dustrial and Labor Relations. Munger, Edwin r g67 A f r i kaner an d A f r i can N a tionalism. L o ndon : O x f o r d Press.

University

Nadel, S. F.

t 956 "T h e C o n cept o f S o c i al E l i t es," I n t e r national Social Science Bulletin 8: 4t g-4s4. Pareto, V. tgg5 The Mind and Society. London: Jonathan Cape. Parsons, Talcott tg5t Th e S o cial System. Glencoe: The F r ee Press. Patterson, Sheila tg57 The Last Trek. London: Routledge and Kegan Paul. Pauw, S. tg46 Di e Be roepsarbeid van d ie Afr ikaner in d i e S tad. Stellenbosch: Pro Ecclesia. Pettigrew, T. F. tg58 " P e rsonality an d S o cio-cultural F actors i n I n t ergroup Attitudes: A Cr o ss-national Comparison." Journal of Conflict Resolution u t 29-49, tg6o "Social Distance Attit udes of South African Students."

Social Forces g8 (March): e46-a53. Pietson, George tg6g T h e Education of American Leaders: Comparative Contnbutions of U.S. Colleges and Universities. New York : Praeger. Pietetse, J. E. (ed.) tg67 V e r slag van Jeugondersoek (4 volumes). Johannesburg: Voor trekkerpers. Porter, J. t g65 T h e V e r tical M o saic. T oronto : U n i v ersity of T o m n t o Press. Rhoodie, N. J. and H. J. Venter tg6o A p a r theid: A S o cio-Historical Exposition of the Origin a nd Development of th e A partheid I dea. Cape Town :

HAUM. Rhoodie, N. J, tg6g A p a r theid and Political Partnership in Southern Africa. Pretoria: Acadetnica.

Rhoodie,N J (ed) rg7z South African Dialogue. New York: M cGraw-Hill. Rogers, Cyril A. and C. Frantz rgfiz R a cial Themes in Southern Rhodesia: The Attitudes and Behavior of the White Population. Neiv Haven and I.ondon: Yale University Press. Sampson, Anthony rgfiz A n atomy of Britain. London: Hodder and Stoughton. Scholrz, G. D. rgb H e t d i e A f rikaanse Volk ' n T o ekomsz Johannesburg: Voortrekkerpers, Bpk. South African Institute of Race Relations rg72 A S u rvey of Race Relations in South Africa, rg7r Jo. hannesburg: S.A. Institute of Race Relations. Statistical Year Book tg66 Statistical Year Book. Pretoria; Bureau of Statistics.

Stiglingh, F. J. rg57 A C o mparative Study of the Public Schools in the Cape Pravince, Unpublished B.Ed. thesis, University of Cape Town. Stultz, Newell rg7s "South African Cabinets and Ministers." South Africa International g (July): r-r8. Thompson, L. M. rg66 Politics in the Republic of South Africa. Boston: Little, Brown and Company. Trapido, Stanley rg63 " P olitical Institutions and Afrikaner Social Structures in the Republic of South Africa." American Political Science

Review g7: 75-87. Trimberger, Ellen K. I g7o "Elite Reinlutions." Sociological Abstracts t 8 (August)-

Suppl. r r: g5a,

Van den Berghe, P. L

rgb South Africa — A Study in Conflict. Middletoivn, Conn,:

Wesleyan University Press. Van der Merwe, C. F. rgfig Di e A f rikaanse Landelike en Stedelike Gesin: 'n Vergelykende Ondersoek. Pretoria: Raad vir Geesteswetenskaplike Navorsing. Van der Merwe, Hendrik W. rg6za "Social Stratification in a C ape Coloured Community." Sociology and Social Research y6 (April): you-gt t. rq6zb "Conflict Among Industrializing Elites in a Saskatcheivan Community." Transactions of the Fifth World Congl.ess $: Q6$-276. rg63 L e adership in a Saskatchewan The I of Industrialization: U C L A d o ctoral dissertation. Ann Arbor Michigan: University Microfilms, Inc.

Commun ity:

m pact

t 966 " C o m munity L e adership i n G ra h amstown." M i m e o graphed repor t. t967 " L e adership in a South African Community." Journal for Social Research t6 (June): t9-35. t97z "Changing Attitudes of Whites Toward African Development." Pp. t o 6 - ti 9 i n C h u rch and Development. Urnpumulo: M i ssiological Institute, t971, Reprint No. ra of the Abe Bailey Institute of I n t erracial Studies. Van der Merwe, Hendrik W., X. C. J. Charton, M. J. Ashley and Bettina J. Huber t97z S t udy of W h ite South African Elites. (Mimeographed report of s t udy submitted t o H u m a n Sciences Research

Council.)

Van der Merwe, Hendrik W. and D, J. Welsh (eds.) tg72 S t udent Perspectives on South Africa. Cape Town: D avid

Philip Publisher.

Van der Nerve, Hendrik W. and J. J. Buitendag t973 " P o l i tical, Ethnic and Structural Differences Among White South A f r i cans." P p . t 8 9 - at 3 in A SS A So c i o logy Southern Africa r978: Papers from the First Congress of the Association for Sociologists in Southern Africa, Durb an: U n i v ersity of N a t al. R eprint 'Ao. t 8 o f t h e A b e Bailey Institute of Interracial Studies, Cape Town.

Van Wyk, S. tg68

Di e Af r i k aner in die Beroepslewe van die Stad. Pretoria en Kaapstad: Academica Pers. Vatcher, William H., Jr. tg65 W h i t e L aager: The Rise of Afrikaner Nationalism. London: Pall Mall Ptess. Vaughan, Michalina t969 " T h e G r a ndes Ecoles" in R u p ert W i l kinson (ed.), Govermng Elites, Studies in T r a i ning and Selection. New York: O x f ord University Press. W arner, W. L l o yd, Paul P. van R i p er, Norman H . M a r t i n a n d Orvis F. Colhns t96g T h e A m erican Federal Executive. New Haven: Yale University Pre s, Weinberg, Ian t967 T h e E n glish Public Schools: The Sociology of Elite Education, New York: Atherton Press. Welsh, David t969 " U r b a n ization and the Solidarity of A f rikaner National-

ism." The Journal of Modern African Studies 7 ( July):

965-276, Whisson, M, G. and Hendrik W. van der Menve i 97z C o l o ured Citizenship in South Africa. Cape Town : Bailey Institute of Interracial Studies. Who's Who of Southern Africa tg68 J o h annesburg: Combined Publishers (Pty.) Ltd.

Abe

t86 Wilkinson, Rupert tg6y T h e P r efects: British Leadership and the Pubhc School Tradition. London: O x ford University Press.

Worrdl, Denis (H.) tg7t

S o u t h A f r i ca : Government and Politics. Pretoria:J. van Schaik.

L.

Zapf, Wolfgang tg65 W a n d lungen der D eutschen Elite : K i n Z i r k u lationmod ell Deutscher Furungsgruppen rgtg-tg6t. M u nchen: R . Piper and Co.

Adam, Heribert, 175, 178 A&ican Homdands, 121, 127 A&icsn National Congress (ANC), 127 A&ican Studies Newsletter, 12 A&ihaner Party, 133 Agriculture, sector, 7 Anatomy of Britain, 113 Anglican (Church), 27, 29 Azqgicisation, 26, 43-5, 37, 86, 173, 176 Anglo Boer War, 115 Archbishop, status of, 96, 97 Armaments Board, 116 Ashley, 59 Assembly, House of, composition and role, 20, 105, 114, 116, 121; education of members, 110; subsector of Legislature, 7 Bachrach, 104 SaltzeH, 3 Banking and Finance, sector, 7 Bantu Authorities Act, 21 Bantu Sdf Government Act, 21 Sasson, Japie, 13$ Beeld, Die> 122 Bendix, 90, 95 Biesheuvel, Simon, 161-3, 174-5 Blau, 12, 13 Blehsley, Dr. A., 134 Sotha, General Louis, 115 Sot tomore, 3 Broederbond, in8uence of, 99, 146-9, 172 Bureaucracy>sector> 7> 8 Burger, Die, 13, 123 Buthelezi, Chief Gatsha> 127

Cabinet, 7, 20, 105, 106 Cabinet Ministers, 23; status of, 95-7 Calpin, 87 Cantril, Hadley, 152 Cape Times, The, 13 Carter, G., 131 Catholic (Roman), 27, 47, 109 C3mnber of M i nes, in8uence of, 146-9 Charton, ct nf, 67 Christian Institute, in8uence of, 146-9 Christian National education, 42 Churches, in8uence of A&ihaans and

English, 146-9

Gild, 12$

Civil Service, national and provincial as subsectors of Bureaucracy, 7, 8 CoHeges, university, technical and teacher training as subsectors of Education, 6 Coloured Persons Representative Council, 22 Cotnmandant-General, status of, 95-7 Communist, N4 Communist Party, 125-7 Company Director, status of, 95-7 Consent, basic element in democracy, 136, 1$7, 149, 17$ Contacts, hdpful in career, 113-4; wtth 0th Y fa ce groups>28> 31>

32 Corporations, Government, 7; public utility, 140 Cronje, 90

Dagbreek, 175 Dagbreek Pers, 128 Dagbreek Public Opinion PoH, 1$4 Defence Force, subsector of Bureaucracy, 7 De ViHiets GraaS; 10$ Dogan, M,, 111 Dutch Reformed Church, and anglicised A&ihaners, 36; and opinions o n legal segregation, 65; a n d support for National Party, 33; membership of fathers of elites, 46-7; percentage of dites and elite language groups, 27W Econotnic dominance, of Englishspeakers, 172 Edinger and Searing> 39 Editor, status of, 95-7 Education, sector, 6; and achievement of dite status, 25, 28 Educational mobility, $0, 31, 80, 81 Elite(s), authority elite, 3; European dites, 6, 12; pohty elites, 104; posidonal elite, 2, 3, 104; power ehte> 2, 3, 104; prestige elite, 3, 104

Ehte cadre, 2, 5, 14 Elite nucleus, 2, 5 Ethnocentricity, in

t h e En g lishspeaking group, 106 Ethnocentrism, of A&ihaans-speakers, 102, 172

Etzioni, 3 Executive dominance, 23, 107 Executive, power of, 105 Fascist, 144 Federasie van Akflraanse Kultuur-

verenigings, influence of, I46-9 Fellow Travdlers, 143 Foreign Intervention, 154, 156, 163 Foreign relations, with Akica, 157-9, 164; with Communist countries, 158-9 Forum, 122 Franz, 65 Free, Lloyd, 152 Frr e Masons, influence rd; 146-9 Future, ezpectations for the, 151 — 164, 172; fears for the, 151-167, 171-2; hopes for the, 151 — 167, 171-2 Gey van Pittius, 137 Gorden, 3, 6, 11-13, 18

Governmental control of individual, 157-9 Grahamstown, 13, 99, 106 Greyling, J. C., 121 Guttman scale, 31, 62, 74 Gut tsman, 58, 111

Hahey, 58, 59 Hammond, 12 Harris, 23, 105, 106 Herrenvolkist, H4 Hertzog, Dr Albert, 130 Hertzog, General J. $. M. (Sarry), 100, 10$, 115, 13$ Hess, 39 Higgins, 109 Homogeneity, of the Akikaans group, 106, 108 Hoofstad, 123 Houghton, D. Hobart, 72, 175 Huber, BettinaJ., ll, 14, 32, 34, 62, 74, 169 Hudmn, William, 161-3, 174-5 Human Sciences Research Wuncii, 11

Hyman, Herbert H., 90, 95 Inequality betwer:n the races, 91-3 Integration,fear of premature, 154,

156, 16$ Interaction, Akikaans-English, 157-9, 164 International Court ofJustice, 127

Jacobs, Gideon, 161-3, 174-5 Janowitz, 5, 39, 79 Jews (Jewish), 27, 47, 108, 109 Judge, status of, 95-7 Judiciary, sector, 7 Keller, Suzanne, 3, 4 Kruger, President Paul„ I 37 Kuper, Hilda, 3 Labour, subsector, 7

Laswell, 3, 39 Legislauue (see also Assembly, house of), 105-7 Leistner, G., 21 Lerner, 3, 6, I I, 12, IS, 18, 39 Lever, 65, 124, 132, 135, 174 Liberal Party, 111, 121, 125, 126 Lipset, 90, 95 Loubser, Jan,86 Louw, I MacCrone, 65

Malan, Dr D. F., 1$3 Malan, Group Captain Sailor, 115 Market Research (Pty) Ltd., 164 Marvick, 110, lll Marxist, 144 M ars Media, sector,6,29 Matanzima, Chief Kaiser, 127 Matthews, 39, 110, 116 Member(s) of Parliament, status of, 95-7; background of, 113 Mills, C. Wright, 3, 6, 53, 58, 115 Moderator of D.R.C,, status of, 95-7 Morris, James 0,, 12 Munger, 35, 86 Nadel, S. F., 3, 104 National Party, and anglicised Akikaner, 36; anti legal segregation,

63-4, 71; andmilitar yservice,115;

percentage of Dutch Reformed Church, SS; percentage of dite language groups, 28-9, 33; policy of, 23-4, 121; structure of, 123 National Union (Party), 133 Nazi, 144 Noodhulpliga, 146 NORC, 95

Oppenhrimer, 103 Opportunities for all, as hopes for the future, 156 Organisation for Akican Unity, 127

18t) Op tion s , sector, 7; subsectors: cultural and entertainment, health and welfare, community service and recreational, 7 Pan A&icanist Congress(PAC), 127 Pareto> 3 Parliatnentary representation, 157-9, 164 Parsons, 122 Patterson, S. C., 110 Patterson, Sheila, 46 Pauw, S., 35, 79 Peace and Harmony, as hope for the future, 156 Pettigrew, 65 Petty Apartheid, 76, 177 Pierson, George, 58 Police, Special Branch of, 11, 12; subsector~ 7 Political alienation, o f E n glishspeakers, 172

Political dominance, of Afrikaansspeakers, 172 Porter, John, 5, 53, 110, 116 Poqo, 127 Press, inliuence of English and Alrikaans, 1~9; political biases of, 13; subsectors: under English control, under Afrikaans control, organisations, 6

Professions, English- and Afrikaansspeakers in, 30; sector, 7, 29 Professor, status of,95-7 Progressive Party, 29, 33, 36, 121, 174; and legal segregation, 64; and military service, 115; policy of, 124-5; structure of, 125 Prohibition of Political Interference Act, 125

Protestantethic,86 Provincial Educational Ordinance, 23 Racial distinctions, index of, 158 Racialist, 144 Radical leftist, 74, 143, 144 Radical rightist, 74 Rapport, 164, 174 Reconstituted National Party (HNP),

122, 123, 130, 132

Red Cross Society, 145 Regression and Discrimination, as fearsfor the future, 156 Research Insdtutes and Museums, subsector, 6

Resistance to elite studies, I 1, 12

Rhoodie, N., 21 Rodgers, 65 Rotary, 146-9 Rousseau, 136 Roux, Professor Ben, 23 Sampson, 113 Scholtz, 86 Schools, Alrikaans, 52 ; E n glishmedlum, 53 g Methodist and Roman Catholic, 55; parallel-medium, 43, 51-2, 111; private, 50-1, 112; private, Anghcan, 54; state (or government), 50-1; state boys, 54 -5; subsector, 6 Sectors, as authority hierarchies in society, 4; delineation of, 8; and representativeness of sample, 14; list of sectors with numbers of eBtes interviewed, 5; and with list of subsectors, 6-7 Segregation index, 160 Self-anchoring Striving Scale, 152-5, 165, 167 Senate, education of members, 110; composition ol; 20, 114, 116; subsector~ 7 Settlers, 1820, 106 Separate Development, ashope for future, 154, 156 Sharpville, 125, 162 Smuts, GeneralJ. C. (Jan), 100, 115, 133 South African Broadcasting Corporation, inguence of, 145-9; nationalisation of, 141; subsector, 6 South African Indian Council, 22 S.A. Statistical Year Book, 1966, 25 Stability, general threat to, as fear for

the fute, 156 Star, The, 13 State President, 20 Stiglingh, F. J., 53, 54 Stultz, Newell M., 116 Subsectors, delineation of, 8; list of, 6-7 Supreme Court, 20 Suzman, H., 100

TEEPS project, 6, 18 Thompson, L. M., 23, 116, 121, 123 Torch Commando, 115

Torney39

IQO Trade Union Council of South A&i~ in8uence of, 146, 148 Transhei Constitution Act, 21 Transvaal Republic Constitution, 137, 138 Travel abroad, aschamcteristic of elites, 27 — 8, 30; and segregation index, 66 Treunucht, 123 Umkonto we Sizwe, 127 United Party, and anglicised A&i%suer, 36; and legal segregation, 64; and military servic, 115; percentage of ehte language groups, 29, 33; policy of, 123-4; structure of,

124 Universities, and education of ehtes, 57-9; and education of polity elites, 113; and soctahsanon of language groups, 44-6; subsectors: English, A&iltaans and bilingual, 6 Uys, Stanley, 122

Van Bgjon, F.J., 141 Van den Berghe, 86 Van der Merwe, C. F., 90

Van der Merwe, st a!., 11 Van der Merwe, Hendrih W., 13, 59, 99, 177, 178 Van Wyh> 79 Vatcher, 102 Vaughan, Michalina,57 Venter, 90 Verhrampte, 38, 122, 144 Verligtes, 38, 122 Verwoerd, Dr H. F., 122 Volts wil, 136 Vorster, J. B., 100 Warner, W. Lloyd, 25 Weinberg, Ian, 53 Welsh, 67 Who's Who of Southern Africa, 14, 15, 16,49, 110 Wilhnson, 53 Wtllq baste element ln

demo cracy

136, 137, 149, 173 World Council of Churches, 171 Worrall, D„21, 22, 105, 106, 121, 122, 139

Zapf, Wolfgang, 3