Abstract
This article is part of a group of Bruinafrikaans (BA) studies, with the primary objective of determining and describing the acoustic quality of its vowels and determining the possibility of a homogeneous nature of this significant variety of Afrikaans. In previous studies strong indications were indeed found in favour of such a characterisation of BA (see Wissing 2022 and Wissing, Baumann and Pienaar 2023.
In the present investigation I focus on the breaking or gliding nature of the mid-high vowel /e/ (referred to as BEES), as reflected in the temporal behaviour of the second vowel formant frequency F2. This is a typical feature that is also present in several Germanic languages, albeit in different ways. For example, in West Germanic Dutch this gliding of F2 is from high to low, whereas in Afrikaans – and unexpectedly also in the North Germanic languages, such as Norwegian and Swedish – the gliding tends to be from low to high. This temporal nature of BEES, as well as the other mid-high vowels /ø/ (REUS) and /o/ (BOOT), can be described in terms of the percentage of the movement of the second vowel formant, F2, from the start of the vowel up to its end.
I use the same research methods as in previous studies on aspects of the BA vowel system mentioned above. The pronunciation of a tried and tested word list, read by 304 speakers of BA, was used to determine the vowel phonemes of the Afrikaans vowel system: /i/, /y/, /e/, /ø/, /ɛ/, /ɑ/, /a/, /ə/, /œ/, /ɔ/, /o/, /u/ in the words kies, nuus, bees, reus, ses, bas, baas, kis, bus, bos, boot, koek, and the diphthongs /əi/, /œy/, /œu/ in byl, uit and oud. The vowels /e/ (BEES) and /ø/ (REUS) are relevant in the current investigation.
The study participants were recruited by trained facilitators in each of the ten areas throughout the Republic (as well as one in Namibia). Recordings of this word list made under the supervision of the facilitators were sent to a central point via smartphone for processing and analysis. The recordings were prepared to extract the appropriate acoustic spectral parameters (F1, F2), which were then statistically processed and used by means of various statistical analyses. A combination of statistical methods of comparison was utilised to analyse and describe the degree of F2 descent in BEES and also REUS to determine the homogeneity within the recording areas. These methods of comparison included determining effect size differences (expressed in terms of d-indices, as proposed by Cohen 1988) regarding the percentage movements of F2, and constructing matrices and dendrograms using standardised Euclidian distance measurements.
The results are varied; they show, to a lesser or greater extent, the homogeneous character of BA. This aligns well with a similar study of the complete vowel system of this variety (Wissing et al. 2023). It indicates that much more work must be done on this important subject. It is a significant possibility that the rounded REUS vowel, due to the widespread presence of the unrounding process in Afrikaans, might be unrounded in BA to such an extent that the phonological contrast between rounded and unrounded vowels is neutralised. The potential merging of REUS and BEES into a single supervowel /e/ is possible. This would effectively double the cases in the BEES dataset, thereby improving the power of the statistical analysis. However, an analysis of the relevant acoustic properties of these two vowels, as found in the speech dataset collected from the ten recording areas, does not provide sufficient findings to support such an amalgamation.
A factor that must be considered here is that all the analyses were done based on the average measurements per area. When individual speakers are examined more closely, one sometimes finds, as in the case of Springbok, deviant cases which are significant in themselves. From a perception point of view, the pronunciation of not only BEES but also the other two mid-high vowels, REUS and BOOT, stand out in such a way that it is quite easy to recognise someone as being from Springbok, something that, according to my observation, is not possible with any of the other nine areas. However, this aspect of the pronunciation of BA as a whole still needs to be investigated in detail.
This also suggests that other speech communities should be considered, particularly in Gauteng and KwaZulu-Natal, which were not included in this investigation. Additionally, the current speech dataset offers an opportunity to learn more about Griqua Afrikaans. I included the recording area of Bergmanshoogte in the Free State, near Philippolis, in this study because it has a significant concentration of coloured Afrikaans speakers who can clearly be identified as of Griqua Afrikaans origin. This variant is generally recognised as one of Afrikaans’ more prominent and important forms. However, an in-depth study of the socio-phonetic characteristics of this variant has not yet been conducted, and such a study would enhance our understanding of Afrikaans in general, and of Bruinafrikaans in particular.
Keywords: acoustic properties; Bruinafrikaans; dendrograms / tree diagrams; d-index; distance matrix; effect sizes; Germanic languages; gliding of F2; gliding percentage; gliding vowels; language varieties; neutralisation; recording areas; second vowel formant; sociophonetics; spectral properties; vowel formant frequencies
- This article’s feature image was created with Canva.