The Serial Verb Construction (SVC) phenomenon is widely researched across many languages. It is generally regarded as a construction in which two or more verbs share the same arguments within a single clause. The verbs in the series must share some grammatical properties such as tense, aspect and polarity. However, there is a verb sequence construction in Dàgáárè that shows apparent similarities to SVCs but with different values for aspect on the verbs. This paper investigates the internal structure of Dàgáárè SVCs and other verb sequence constructions such as multi-aspectual constructions (MACs) and coordinate structures. Applying a variety of syntactic and semantic tests, the paper distinguishes SVCs from MACs and coordination and shows the relation between MACs and coordination. Based on the results of the tests, I argue that although MACs have some properties of SVCs, they are not SVCs. Rather; I conclude that MACs pattern with coordination or covert coordination in Dàgáárè and they are perceived to express distinct events.
In Jumjum, a Western Nilotic language, some body-part nouns, and only such nouns, may be externally possessed in transitive and antipassive clauses. In these external possessor constructions, the possessor is either the object of a transitive verb or the demoted patient of an antipassive verb. The externally possessed body-part noun is partly incorporated into the verb, as shown by the following properties: It immediately follows the verb, its tone is determined by the final tone of the verb, it may combine with a nominalized verb in a kind of compound, and it does not exhibit the root-final nasalization that is prevalent in monosyllabic singular nouns in Jumjum, including internally possessed body-part nouns.
This paper uses historical-comparative approaches in combination with quantitative methods to analyse data from a survey of varieties of the Bantu languages Herero and Kuvale spoken by ethnically diverse groups from southwestern Angola. We assess the status and position of the underdocumented “Kuvale” variety in relation to its closest geographic neighbours, and address questions about the history of the area. We find that Kuvale is lexically differentiated from its closest relatives Herero, Wambo and Nyaneka-Nkhumbi and should probably be considered a language in its own right. Within the lexicon and phoneme inventories of the surveyed varieties, no obvious indications of a substrate were found, including in data collected among the formerly Kwadi-speaking Kwepe, and among the Kwisi and Twa foragers, who have been hypothesized to constitute a remnant layer of non-Bantu, non-Khoisan foragers in the Namib desert.