The role of pause in discourse and its place in linguistics: Some evidence from Eastern Bantu

Abstract

Abstract Relatively few studies of breaks in the flow of speech assume a role in linguistic structure for pause. Using evidence from pause placement in several Eastern Bantu languages, including Kikuyu (from which all examples cited in the paper are drawn), it is possible to show that pause is not random, nor physiologically conditioned, nor explicable in terms of planning speech-bursts to come. There is a consistent contrast of three lengths of pause, with a ratio 1:2:3 or 1:2:4. The shortest pauses occur, apparently optionally, at word, phrase, or clause boundaries. The two longer categories, however, correlate with linguistic boundaries at the discourse level. The pattern of three pause levels is found in all the Eastern Bantu languages so far investigated, as a feature of narrative. If cases of hesitation are carefully distinguished from significant pause, it can be shown that the same pause hierarchy is part of normal conversation, at least for Kikuyu and other Dhaagiew languages. It is concluded that pause should indeed be seen as a linguistic category, though other languages have yet to be investigated, the possibility remains that pause is universal and automatic, rather than language-specific and linguistically significant. However, the fact that Kikuyu and Cinyanja differ, even though minimally, with ratios of 1:2:3 and 1:2:4, respectively, encourages acceptance of the pause as a valid linguistic unit.