This paper assesses the extent to which the relative system of Reunion Creole can be considered evidence for the "semi-creole" status of the language, which Holm (1989, p. 392) has attributed to it. As a "semi-creole", Reunion Creole is often left out of creole grammars such as that by Syea (2017). This pilot study therefore begins to fill a typological gap in the literature, outlining the strategies used in headed relative clauses in the language. It compares these strategies against those used in French (its lexifier) and the other Indian Ocean Creoles. For subject and direct object relatives, it finds the relativising strategies closer to those of the "full" creoles of the Indian Ocean. In these positions, Reunion Creole employs a zero and gap or relativiser and gap strategy, at rates of roughly 80% and 20% respectively. The optional relativiser used is universal ke. Oblique relatives prefer a relativiser and gap strategy, and an additional relativiser, ousa, is found in locatives. These relatives thus exhibit some more similarity to French. Genitive relatives are found to prefer zero-marking and a pronoun retention strategy, therefore more closely resemble the Indian Ocean Creoles. However, the study also finds some evidence of a relative pronoun strategy with obliques and genitives. It thus highlights the need for further research to establish whether this strategy is in wider use, and whether this possible development of a relative pronoun strategy is an internal development or a result of language contact with French. If further research finds a relative pronoun strategy in wider use, it may support the "semi-creole" classification of Reunion Creole. However, the results of the present study suggest that the relative system of Reunion Creole resembles more closely that of the "full" creoles of the Indian Ocean than that of French.

Background: The current study examined the effects of depression on vowel production and vowel space area (VSA). Depressed speech was compared to that of healthy controls to determine whether vowels differed quantitatively according to the first (F1) and second (F2) formants. VSA measures were also analysed in relative to depression severity to determine whether these were clinically meaningful indicators of affective state. Research design and methods: Vowels were extracted from the read speech of nine depressed and nine healthy participants over four recording time points. Raw formant values were normalised and plotted to compare group means. VSA measures were examined in relation to depression severity using Kendall's tau. The predictive value of diagnostic group for formant measurements (F1 and F2) were assessed using linear mixed effect models. Results: Variable differences emerged for vowel types between depressed and healthy participants. Significant interactions emerged for group and vowel type, with groupwise mean comparisons indicating that low vowels /æ/ and /ɑ/ are disproportionately affected by depressive status. Global VSA measures were not significantly correlated with depression severity scores due to these asymmetric effects. Discussion and implications: Specific vowels may be linked to depressive diagnosis based on a combination of motor incoordination and facial tension, particularly with respect to jaw and tongue height. Global measures of VSA were not sensitive indicators of depressive symptom severity.

As part of a larger PhD study entitled A comparative phonological analysis of varieties of English spoken by native speakers of Nigerian languages - Hausa, Igbo, Kanuri & Yoruba - for the determination of speakers' origins, recordings were made of speakers from Hausa, Igbo, Kanuri and Yoruba groups in Nigeria. Four different styles of speech were represented in English: spontaneous speech, picture descriptions, reading of phonetically-balanced passages and that of a word list. From 15 participants in each of the four groups, three participants were selected from the Hausa and Kanuri groups for the initial work. Their speech samples have been analysed at the segmental phonological and phonetic levels for features that distinguish one variety from the other. The paper aims to answer the following questions: (1) In what ways does Standard Nigerian English spoken by L1 Hausa speakers differ from that spoken by L1 Kanuri speakers? (2) Can the differences be accounted for by reference to the different L1 phonological systems? It is hypothesised that the Hausa and Kanuri English varieties are differentiable, and that major disparities between them, attributable to L1 phonological interference, will be apparent in their segmental phonological and phonetic features. The latter provide the main focus of this paper. Due to space restrictions of the article, only realisations of five English fricative consonants, as well as the velar nasal and voiceless bilabial plosive will be reported here. As evident in the findings of the study, all five of the English fricatives were pronounced differently than in British English by both Hausa and Kanuri English speakers. Though some of their realisations sound the same, they tend to differ in several ways; some of which are determined by the sound's environment, speech style (free speech or reading) and absence of the target sounds in the speakers' L1.