The effects of language on the informativeness and accuracy of child witnesses in Kenya

Kariuki, Moses

The effects of language on the informativeness and accuracy of child witnesses in Kenya

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Department of Psychology

Faculty

Faculty of Humanities

Abstract

This report describes the first prospective study that we are aware of specifically designed to assess the effects of language on children's informativeness and accuracy for a staged memory event one week after it occurred. The study took place in a school which is located in rural Central Kenya where Kikuyu is the dominant home language for the locals. Hence, all seventy-six 9-15-year-old Grade five pupils who took part in the study spoke Kikuyu as their home language. They also learnt English as a subject and used it as a language of instruction and examination for all the other subjects. However, Kiswahili is taught and examined as a standalone subject. The language of the memory event was Kiswahili. Half of the participants were interviewed about the memory event in their home language (Kikuyu) while the rest were interviewed in English, an official language of the court. The children who were interviewed in Kikuyu were significantly more informative and accurate about the event than those who were interviewed in English. The findings from this study are presented together with (and supported by) reports of four studies. First, through an analysis of the legal framework supporting child witnesses in Kenya, I found that children are supported by an impressive and progressive wide-ranging legal framework, recognising both home-grown and international instruments to which Kenya is a signatory. However, there is need for studies to establish whether these special measures are well implemented, and that they achieve what they are meant to achieve. Also, from a review of case law, it emerged that important language-related concerns exist. In addition, magistrates and prosecutors reported through a survey that a significant number of child witnesses in Kenyan courts either have some challenges communicating adequately in the language of the court or are totally unable to communicate in the language of the court. Finally, an analysis of actual court transcript of a child witness under examination-in chief and cross-examination presents some evidence that child witnesses in Kenya could be enduring poor interviewing practices. This thesis provides some credible evidence that child witnesses' home language should always take precedence when questioned in court or during investigations by police interlocutors. These findings have important empirical and applied implication in the field of forensic interviewing.