Volume 32, Issue 2

In Mbuun (B87), a Bantu language spoken in the Bandundu Province of the Democratic Republic of the Congo, the valency-increasing functions of the causative and the applicative have, to a certain extent, the same formal expression. They are encoded morphologically by a derivational suffix which has an uncharacteristic morphophonological realization. It triggers the gemination of the root-final consonant. In this paper, we examine whether this causative/applicative isomorphism is the result of a semantic split or rather the outcome of a phonemic merger. In the first case, the semantic extension must have happened from applicative to causative rather than the cross-linguistically more common evolution from causative to applicative, since the consonant gemination associated with these forms has its origin in the Proto-Bantu applicative suffix * - i d- . In the case of a phonemic merger, however, this applicative suffix would have become homophonous with an originally distinct suffix having a causative effect as the result of convergent morphophonological change. This morpheme cannot be the Proto-Bantu causative suffix * -ici- , which still has a residual causative use in Mbuun.

This paper investigates the distribution of argument and adjunct question phrases in Ikalanga. It is argued that both adjunct and argument question phrases in this language are focused, but that the former are merged in the specifier of a focus phrase (which projects either within v P or higher than TP) while argument question phrases occur in the specifier of a focus phrase as a result of a syntactic operation, namely attract . Evidence that adjunct question phrases are base generated in spec-FocP rather than moved there while argument question phrases are moved to spec-FocP comes from morphological differences observed between the two construction types as well as movement tests. There are two main points of interest related to the distribution of adjunct question phrases. One has to do with the fact that Ikalanga has one word chini which means ‘how’ or ‘how come’. The paper argues that speakers distinguish chini ‘how’ and ‘how come’ based on their syntactic distribution, which is complementary. chini ‘how come’ is always merged in the specifier of FocP and can occur either on the left edge of the sentence or on the right edge of the sentence. chini ‘how’ on the other hand only occurs at the right edge of the sentence or between the verb and its complement. It is shown that although chini ‘how come’ seemingly occurs on the right edge of the sentence, this is a derived position: it is a position that results after a proposition, a TP, has been topicalized leaving behind chini in spec-FocP. The second point of interest concerning the distribution of adjuncts is the rather unusual position they occupy, namely between the verb and its complement. The paper proposes that these data can be analyzed successfully by adopting the proposal that a focus phrase projects within v P (see Ndayiregige, Linguistic Inquiry 30: 399–444, 1999, van der Wal, The disjoint verb form and an empty immediate after verb position, ZAS, 2006, among others, for similar proposals.

Mande languages are characterized by a typologically unusual rigid S-O-V-X word order pattern, and do not fit neatly into the traditional category of “verb-final” languages. While subjects and objects precede the verb, all oblique arguments and adjuncts follow it. Several diachronic hypotheses have been advanced to account for the word order type of Mande (Givón, Serial verbs and syntactic change: Niger-Congo, University of Texas Press, 1975, Heine and Reh, Grammaticalization and reanalysis in African languages, Helmut Buske, 1984, Claudi, Die Stellung von Verb und Objekt in Niger-Kongo-Sprachen: Ein Beitrag zur Rekonstruktion historischer Syntax, Universität zu Köln, 1993), none of which are fully satisfactory. This paper discusses new evidence from the syntax of modern Mande languages that may shed light on the historical development of the S-O-V-X word order. I argue, based on the unusual behavior of Mande postpositional phrases, that the development of S-O-V-X was a consequence of categorial reanalysis of constructions with deverbal nouns as verb phrases, with subsequent replacement of the older type of verb phrase by the newly introduced structure. The analysis is supported by a number of characteristic parallels between the syntax of Mande verb phrases and noun phrases, such as the absence, in some languages, of both verb-phrase-internal and noun-phrase-internal postpositional phrases, the absence of the genitive vs. accusative distinction, and identical restrictions on the number of complements of nouns and verbs.

An analysis of the segmental phonology of Koorete ( kòorete ) is presented, with sections on vowels, consonants, and consonant clusters. The analysis differs from earlier ones (Hayward, Afrika und Übersee LXV: 211–268, 1982, Ford, Notes on Ko:rete phonology, School of Oriental and African Studies, University of London, 1990) by claiming that diphthongs constitute an important part of Koorete phonology, that Koorete has no distinctive opposition between affricates and fricatives, and that there are distinctive oppositions between single and geminated sibilants in consonant clusters. There are also shorter sections on aspects of Koorete tonology and morphophonology.