This article analyzes stereotypes of Muslims that have recurred in Sinhala literature over the past seven centuries. This temporal span includes (1) a precolonial time when Muslims were a curiosity in Sinhala poetry as rich traders, wild men, and seductive women; (2) an early colonial time when the Portuguese and Dutch displaced more Muslims into the island interior, and Sinhala authors increasingly wrote of them as religious others akin to Tamils; and (3) a late colonial time when British policies forged religo-racial political categories in the decades leading up to the anti-Muslim pogroms of 1915. Each case is also connected to postcolonial instantiations or transformations of these typecasts. This history therefore eschews linear narratives of change to show the recurrent tendencies of social reasoning through stereotyping, past and present.

We report the process of adaptation into Sinhala of a questionnaire given to mothers of twins to determine zygosity. Adaptation and validation was carried out in three stages. Firstly, we used a nominal group to translate the English version and to assess the extent of agreement (consensus measurement) on the appropriateness of the translation and resolve disagreement (consensus development). Secondly we used a qualitative interview with 25 mothers of twins. The three main stems of the translated questionnaire were used as a semi-structured interview, and the responses noted verbatim. These were categorised and analysed, and the translated full questionnaire was then presented as closed questions with fixed choice responses. The categorised responses generated during the qualitative interview were compared with the responses to the fixed choices in the full questionnaire. The third stage was the appraisal of the questionnaire by 17 bilingual parents of twins. The source and translated version of the questionnaire were given to them at least 3 days apart. The responses were rated and the total scores were computed to determine the zygosity. This step was carried out to measure the validity and reliability of the Sinhala version. A perfect correlation between the original and adapted version was obtained, with a kappa of 1. The results suggest that the Sinhala version of the questionnaire is conceptually equivalent to the original questionnaire. Comparison of the zygosity determination by using this adapted questionnaire with results from analysis of genetic markers on Sri Lankan twins is needed for final validation of the translated questionnaire. Twin Research (2000) 3, 205–212.