Volume 40, Issue 1 -

Rutooro, a Bantu language spoken in western Uganda, exhibits two processes common to Bantu languages: Mid Vowel Harmony (MVH), by which a high vowel is lowered to mid if it follows a mid root vowel, and Consonant Mutation (CM), by which a high front vowel of certain morphemes trigger a change in the immediately preceding consonant. In forms where both rules could potentially apply, CM applies, but MVH does not, even though its structural description is met. Ultimately, it is shown that MVH is blocked when an alveolar fricative intervenes between the trigger and the target. The most anomalous forms are those that contain the short causative / -i / and the Perfective suffix / -ir /. In some forms CM would be expected to apply, but does not, and in other forms MVH is expected to be blocked, but is not. The key to the analysis lies in the fact that CM is not a purely phonological process, but rather is only triggered by a small set of specific morphemes. The surface anomalies in the perfect forms can all be accounted for if one formalizes the relevant processes as only being triggered (or blocked) by unaltered [ -ir ].

The systematic comparison of the different types of progressive Vowel Height Harmony (pVHH) attested within the Kikongo Language Cluster (KLC) leads to the conclusion that this common Bantu process of long-distance assimilation cannot be reconstructed to Proto-Kikongo. The ‘(a)symmetric-pVHH’ and ‘back-pVHH’ patterns, the two main and structurally different kinds of pVHH within the KLC, emerged independently and relatively late within two distinct subgroups, viz. South Kikongo and North Kikongo respectively. Moreover, the ‘(a)symmetric-pVHH’ pattern further spread from a South Kikongo focal area coinciding with the heartland of the Kongo kingdom to other parts of the KLC through contact-induced dialectal diffusion. Furthermore, the historical-comparative evidence from the KLC suggests that neither symmetric nor asymmetric pVHH should be reconstructed to Proto-Bantu, the most recent common ancestor of all Bantu languages.

This paper provides an overview of noun class systems in a sample of 20 Kwa languages. It focuses on the synchronic productivity of noun classification in Kwa as opposed to the full-fledged class system assumed for Proto-Kwa and for the general “Niger-Congo prototype” (Good, Jeff. 2012. How to become a “Kwa” noun.” Morphology 22(2). 293–335; Creissels, Denis. Forthcoming. Noun class systems in Atlantic languages. To appear. In Friederike Lüpke (ed.), The Oxford guide to the Atlantic languages of West Africa . Oxford: Oxford University Press). The productivity of class morphology on nouns is studied by exploring class marker alternations as exponents of other grammatical phenomena: the formation of the plural, diminutive derivation and nominalization. We also discuss class indexation on nominal modifiers (adjectives, numerals, demonstratives and indefinite markers) as well as subject and object pronominals. We demonstrate that Kwa languages tend to follow the typological tendencies pertaining to class marking on nouns and class indexation, e.g. the Agreement hierarchy (Corbett, Greville G. 1979. The agreement hierarchy. Journal of Linguistics 15. 203–224), as established for world languages in general and Niger-Congo family in particular (Good, Jeff. 2012. How to become a “Kwa” noun.” Morphology 22(2). 293–335). However, some intriguing discrepancies, e.g. different patterning of indexation on adjectives vs. on numerals in Kwa as opposed to some other Niger-Congo branches, were also attested. In diachronic perspective, our findings suggest that noun class systems are flexible as they show high intragenetic variation and are easily degradable, but they almost never disappear completely.

In this article we present comparative data on visual hunting signals from the Kalahari Basin Area of southern Africa, encompassing three Kalahari Khoe-speaking (Ts’ixa, Buga, ǁAni) and one Kx’a-speaking group (Juǀ’hoan). For the comparison, an analysis of handshapes, handedness and iconicity in the individual data sets was conducted. Being applied in analyses of gestures and (partly) sign languages, these parameters reveal significant similarities and differences between the individual systems. We find that the systems documented most likely constitute an areal feature that transcends the boundaries of genealogical linguistic affiliation and may ultimately contribute to a better understanding of population contact and local networks in the Kalahari Basin Area Sprachbund .