Referring to information source in Kirundi: The case of lexical evidential markers.

Abstract

The aim of the present study was to investigate ‘evidentiality’ or referring to information source in Kirundi, through an appraisal of the evidential function and meaning of four invariable expressions, namely ngo ‘it is said that’, ngira ngo ‘I think that’, umeengo ‘apparently’, and usaanga ‘it appears that’. The theoretical approach on which the research was based is that evidentiality is a semantic-functional domain, which is expressed in languages by either grammatical or lexical markers. Willett’s (1988) classification of evidential types is the analytic model used by the researcher to analyze and present the findings. The data were collected from a variety of sources including natural conversations, recorded interviews, and written texts such as newspaper items and reports of different types. The simple random sampling method was used to collect and sample data.

The findings of the study are of two kinds. First, the evidential device ngo ‘it is said that’ is used to mark reportative evidentiality in Kirundi, where second-hand evidence, third-hand evidence, and folklore evidence are represented. Second, the expressions ngira ngo ‘I think that’, umeengo ‘apparently’, and usaanga ‘it appears that’ are markers of inferential evidentiality; and the inferred evidence is based on either result or reasoning. These evidentials are used with some distancing effect to achieve hedging, hence marking inferential evidentiality with epistemic overtones.

The contribution of the study is the validation of the classificational model of evidentiality suggested by Willett (1988), especially concerning the subdivision of evidential values within the reportative domain of indirect evidentiality. The findings of the analysis of Kirundi data confirm folklore evidence, which is subcategorized within reported evidence and which does not appear in a number of other classifications of evidential types. Therefore, the present study of lexical markers of evidentiality in Kirundi confirms the relevancy of studying the lexical expression of evidentiality since such studies contribute to “determining a consistent classificational model of evidentiality as a linguistic category” (Squartini, 2008:941).