Lexical processing: Exploring the use of cognates by multilingual speakers

Abstract

This paper studied cognates as an aspect of lexical processing among Burundian multilingual speakers. Thirty people who could speak Kirundi, French, English and Kiswahili participated in language production through an oral semi-structured interview. The aim of this paper was to investigate the types of cognates produced by Burundian multilingual speakers. To achieve this aim thirty (30) people participated in the interview. The language produced by participants was transcribed and utterances containing cognates were highlighted and analysed through themes building. It was found that among cognates produced by participants, there were true cognates from French and English. Their main difference is pronunciation. Participants pronounced these true cognates as they would be pronounced in French, therefore making them inappropriately pronounced in English. In addition, false cognates were also produced and they were associated with words from either French or English. Words which are highly identical in both forms and meaning in the two different but similar languages were simultaneously activated. Surprisingly there was no production of cognates from Kiswahili and Kirundi even though these languages are very close (both being bantu languages). It was concluded that speakers of both French (where French was learned before English) easily access words in French and in case the same words exist in English, the words are simply used as such. This leads to incorrect use of such words in the target language (English in this study). Besides, processing English words leads to production of false cognates which are considered to be appropriate but which are actually wrong in English. It was recommended that since both languages involved in the cognates production are learned as foreign languages, the teaching of these languages should systematically consider the existence of the cognates in a purposive scaffolding.