The Zanzibaris of Durban constitute the smallest minority in South Africa's ethnically diverse society. The largest cluster of Zanzibaris reside in a predominantly Indian area of Bayview, Chatsworth. Their adolescents attend schools with black and Indian peers. The theoretical perspectives of self-esteem and social distance imply that adolescents in such circumstances face a complex task of identity formation. This study compared levels of self-esteem, as measured by the Piers Harris Self Concept Scale, in Zanzibaris, Zulu-speaking blacks and Indian boys and girls aged 13-16 years. The sample consisted of 263 respondents of 3 racial groups (Zanzibaris (n=60); Indians (n=154) and Zulu-speaking blacks (n=49), of both genders drawn from two urban schools in Bayview, Chatsworth (a socio-economically heterogeneous area) in Durban. An adaptation of the Bogardus Social Distance Scale was employed to assess the attitudes of Zanzibari adolescents to other racial groups. These attitudes were examined for gender differences and in relation to self-esteem scores. In view of the sensitivity of the study, parental consent was sought and respondents were briefed before and after administration of the measures. The results were analysed using analysis of variance, t-tests and correlation co-efficients. Interracial comparisons did not uphold the prediction that Zanzibaris would exhibit significantly lower levels of global self-esteem or its six components. Zanzibaris displayed significantly higher scores globally and for five of the components. No significant gender differences were found in self-esteem scores of the entire sample or for each racial group. The prediction, that Zanzibari adolescents in view of their circumstances, would show greater social distance towards blacks than towards Indians was not supported. Zanzibari boys and girls were similar in their ranking of other racial groups in terms of out-group preference, with boys showing greater social distance towards each group. Self-esteem and social distance scores were positively correlated at a non-significant level. This did not support the prediction that minority adolescents who preferred out-groups over in-groups would have lower self-concept scores. The findings are discussed in terms of theories of self-esteem, social identity and contact hypotheses, and contrasted, with those of other studies conducted in South Africa and abroad. Attention is drawn to the strengths and limitations of this study. The findings have implications for policy makers at the level of school and community in order to reduce prejudice and promote intergroup harmony. It is suggested that curriculum packages include social science sessions to explore concepts of tolerance, racism and inter-ethnic communication both at individual and institutional levels. In the light of this study, suggestions are made for further research to inform the discourse around marginalised minorities.

Thesis (M.Ed.)-University of Natal, 1998.