Previous studies on Gizey (glottocode: gize1234, Masa < Chadic < Afroasiatic) have treated the monosyndetic and bisyndetic uses of the conjunction māj/mɛ̄j as semantically equivalent. This paper challenges that view, presenting novel evidence that these two forms of conjunction cannot always be equated. We argue that the distributional differences between monosyndetic and bisyndetic conjunction can be better understood by considering how they contribute to the coherence of the discourse that contain them. Using Segmented Discourse Representation Theory (SDRT), we show that monosyndetic and bisyndetic conjunction are compatible with coordinating discourse relations (e.g. RESULT, NARRATION, PARALLEL, CONTRAST) and are congruently incompatible with subordinating discourse relations (e.g. ELABORATION, EXPLANATION). Furthermore, whereas monosyndetic conjunction can appear with all coordinating relations, bisyndetic conjunction is restricted to the structural discourse relations, namely PARALLEL and CONTRAST.
Continued in-depth historical-comparative research into Central Chadic phonology reveals that present-day Central Chadic languages possess a number of phonemes that cannot be reconstructed for Proto-Central Chadic. Among the driving forces for their evolution were processes of ‘prosodification’, whereby articulatory features such as PAL (palatal), LAB (labial), NAS (nasal) and GLOT (glottal) disassociate from their original host segments and become ‘floating’ and thereby free to re-associate with other hosts within the word. In Chadic linguistics, such suprasegmental processes and effects are referred to as ‘prosodies’. As a result, present-day languages possess innovative palatalised, labialised, prenasalised, and glottalised consonants and approximants as much as considerably enriched vowel systems. The paper discusses diachronic scenarios according to which ‘secondary’ glottal(ised) segments emerge in addition to safely reconstructed ‘primary’ *ɓ and *ɗ. In one scenario, allophonic glottal stops end up as a floating feature (GLOT), which re-associates with another segment in the chain of phonemes resulting in glottal ‘colouring’ of the new host. In a parallel scenario, an allophonic glottal stop ends up in abutting position to another segment, with which it fuses and likewise fosters a new synchronic glottal phoneme.
The typological generalizations that emerge from comparing phoneme inventories and phonological processes in as many of the world’s languages as possible are the foundation of theory building in phonetics and phonology. Since one-third of the world’s languages are spoken in Africa, one would expect them to be well-represented in typological surveys of segment types and segment inventories. However, this is not the case. This paper aims to address that gap by surveying, specifically, the occurrence of laryngeal stop qualities – voiceless unaspirated, voiced, voiceless aspirated, implosive and ejective – in a large (300+) geographically and genetically diverse sample of African languages. The results of the survey turn out to challenge common assumptions about laryngeal markedness for stops and stop inventories, demonstrating the importance of this kind of areally-focused typological survey and the need for more phonetic investigations of African languages as the basis for future surveys.
In Nuer, a highly fusional and predominantly monosyllabic West Nilotic language, morphology is expressed primarily by stem-internal changes that affect vowel quality, voice quality, tone, vowel length and the stem-final consonant. This paper focuses on the interrelated alternations in vowel quality and voice quality, showing how these two parameters are used to encode inflection and derivation. Because the relation between the sets of the participating vowels is not systematic in a phonological sense, the alternations are analysed by postulating abstract morphological categories termed vowel grades. The proposed analysis enables us to unambiguously categorise the patterns of morphological marking on verbs, nouns, and adjectives. In addition, it simplifies the vowel inventory relative to earlier accounts. Finally, the proposed analysis contributes to our understanding of the diachronic development of vocalic phenomena in the West Nilotic languages.
Previous work has established that focus in Luganda is clearly signaled through the choice of syntactic configuration and morphological marking. There has been no indication in the literature to date that there is any use in the language of acoustic prominence cues for focus such as those found in English or Mandarin.The present article provides experimental evidence that vowels in focused expressions in Luganda have greater duration, higher peak f0, greater peak intensity, and more compressed postfocal f0 range than corresponding vowels in nonfocused expressions. Such acoustic marking of focus is largely redundant in Luganda, due to the clear syntactic and morphological marking, but such redundancy is an effective adaptation to a noisy environment.On the basis of the combined cues from the syntax, morphology and phonetics, Luganda speakers were able to accurately match a statement to the Wh-question that it answers. Allowed a free choice of answering strategies, Luganda speakers displayed a strong preference for the preverbal focus construction (a cleft-like structure) and for placement of the focused phrase at the beginning of the answer statement.
