Spurious high-tone extensions in Luganda

Abstract

A handful of mostly older descriptions of Bantu tone systems note an unexpected H tone effect on verb stems containing a causative -j- or passive -u- extension. Meeussen cites these effects as archaic and reconstructs H tone on these vocalic suffixes in Proto-Bantu. It has been difficult to evaluate this historical claim, since not all Bantu tone systems exhibit what we are calling ‘spurious H-tone extensions’, and still fewer of these have been described in detail. In this article we present a reasonably complete description of spurious H-tone effects in Luganda, demonstrating that spurious H can (and must) appear only when these extensions occur on a ‘modified base’ (<PB *-jd-e) verb form in an inflectional frame that independently requires a H tone suffix. We suggest that the unexpected additional H tone is not underlyingly on the vocalic suffixes, but rather has the status of an enclitic in synchronic Luganda. We hypothesize that spurious H is a relic of an earlier stage of Bantu when the verb stem had not yet been completely forged into a single unit. This allows us to provide a diachronic account both of its present-day distribution and of two other effects its presence has on the verb: obligatory final -a instead of the more usual -e of the modified base, and blocking of tone-group formation when it would otherwise occur.