• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
Vowel derounding

Derounding of the marked rounded vowels is one of the most salient characteristics of Afrikaans. Age and gender are important factors in explaining differences with respect to the scale of derounding.


The derounding of Afrikaans marked rounded vowels has been in evidence since as early as the 18th century (Ponelis 1990). Le Roux and Pienaar (1927) mention this phenomenon in strongly negative terms: “plat … en sterk af te keur” (p.58) ( unrefined ... and to be avoided strongly). In contrast to this, De Villiers and Ponelis (1987) declare derounding in both koiné and Cape Vernacular Afrikaans as a given. Wissing (2010) states that derounding is a process that is an important distinguisher between Afrikaans on the one hand and the other Germanic languages, such as Dutch and Swedish, on the other; and adds that, even should derounding not be present, the difference between the rounded Afrikaans vowels and their Dutch and Swedish counterparts is notable: the acoustic distance between the rounded and unrounded vowels of these two Germanic languages is much greater than in Afrikaans.

Initially, the result of this process was thus viewed in strongly negative terms as being substandard; more recently, the presence thereof is described in neutral terms as a typical feature of Afrikaans and impacting on a number of vowels: thus the unrounding of /y/, /ø/ and /œ/ to resp. [i], [e] and [ə], as well as of the diphthong /œy/ to [əi], is among the most distinctive features of Afrikaans pronunciation compared to other Germanic languages. Derounding is most observable in casual speech, but is, by far, not restricted to this style. Even in formal language, as spoken by radio and television personalities, or on public occasions such as political speeches or sermons, it is readily heard. Lack of stress, as well as unaccentedness in connected speech, is conducive to, but not limited to, this phenomenon. Only in cases of extremely unknown and thus seldomly used words does the chance of derounding diminish.

Production and perception studies by Wissing(1994), (1995), Wissing and Van Dijkhorst(2006), Wissing(2011) and Wissing(2017) as well as further, unpublished investigations, all reveal the following picture of derounding among speakers of different genders and ages:

  • Older speakers, especially those who are 70 years and older, do not deround as much as young speakers do.
  • Younger speakers have considerably less difficulty in a discrimination task in which rounded ~ unrounded word pairs such as kleure colours and klere clothing are involved. In the relevant experiment such pairs were produced by both older speakers and younger speakers. In the latter case a RAP of only 0.43 was achieved; for productions by the older speakers nearly 1.0 was registered. These results point in the direction of the existence of two chronolects. This, in turn, could also be evidence for a diachronic shift in the case of the vowel system of Afrikaans. Compare also the section Age as a factor lower down for more specific information on this matter.
  • Although derounding is present over a broad spectrum of modern Afrikaans, absolute neutralization of the contrast rounded ~ unrounded for these marked vowels has not yet taken place .

Some recent unpublished investigations into vowel derounding underscores the impression that derounding in Afrikaans is indeed a wide-spread phenomenon (For speech data sets that were used in these investigations, see Introduction to phonological processes).

In the following sections some of the most interesting findings from the analysis of these data sets regarding vowel derounding is presented.


Vowel derounding is a special type of phonological process in the sense that it is not particularly sensitive to phonetic context, that is, it commonly takes place even in monosyllabic words, thus not specifically in unstressed positions of multisyllabic words. An example of the former is the word nuus /nys/ news, pronounced in its unrounded form as [nis]. Of course, the known factors which enhance the application of phonological processes (see Introduction to phonological processes) are still relevant here. Therefore, for example, the rounded vowel /y/ in well-known and frequently used words like minuut /mənyt/ minute, uur /yr/ hour and musiek /mysik/ music is more likely to be unrounded than in less well-known and less-frequently used words such as immuun /imyn/ immune and vakuum /vɑkym/ vacuum. Speech style and rate are other contributing factors.

In the following sections a systematic description of derounding against this background is presented.

[+]Tempo as dominant factor

Wissing and Van Dijkhorst (2006) found strong evidence in favour of the view that the RAP of derounding is governed by speech rate, both in a reading task (a passage of prose) and in spontaneous speech (story telling). The three candidates for potential derounding were /y/, /ø/ and /œ/, as in, for example, the stimulus words uur hour, neus nose and bus bus. Derounding resulted in resp. [i], [iə] and [ə]. Two speaking modes were controlled for, rendering results for NormTemp and FastTemp (resp. referred to elsewhere as RAN and RAF). The results are indicative of a greater tendency for derounding to take place at a fast speaking rate (an RAP of 0.65 for RAF compared to 0.57 for RAN).

[+]Frequency of usage / familiarity of words

It is clear that derounding as a phonological process is sensitive to these factors. In the RSG data set, supposedly read in a formal albeit relaxed style, the function word deur /dør/ by, for example, occurs 267 times; of these, 76 were rounded, that is, a RAP of 0.7, which is very much in line with the RAP of the FastTemp mentioned in the previous section. Being a function word, deur is one of the most frequently used Afrikaans words. In contrast, the vowel of nuus, in the same data set, had an RAP of only 0.25 for derounding of (i.e. 4 out of 20 possible cases). This relative proportion of 267: 20 tokens of deur: nuus is in accordance with the number of occurrences of these two words in VivA's corpus of nearly 85 million words. In this corpus, deur appears 11777 times, compared to only 64 times for nuus. These findings strongly support the general assumption that frequency of usage is an important factor in predicting the application of phonological rules. Similarly, the probability of derounding, from /œy/ to [əi], in the case of the diphthong in the frequently used function word uit, is significantly higher than that in much less used and less known content words like snuit /snœyt/ snout or kruis /krœys/ cross.

A special case is that of place names ending in -burg /bœrx/ e.g. Johannesburg, Rustenburg, Sasolburg, Pietermaritzburg, Postmasburg, Vryburg, Vredenburg, all names occurring frequently every day during the Afrikaans national radio broadcaster, RSG's, weather reports. An inspection of the reading by one prominent RSG-presenter revealed that the realization of the u-vowel (i.e. the /œ/ phoneme) in these place names, but also in words like temperature temperature and minus id., is almost almost always unrounded i.e. [ə]. Words ending on -or behaves somewhat differently. On the one hand, derounding of the vowel in /ɔr/ regularly produces a [ə] as in doktor, motor, professor, junior and senior, but not in faktor, lektor, rektor and sektor – all loan words from English, and all of them with more or less equal frequency usage. doktor is especially interesting in that dokter – a physician – and doktor – a person holding a PhD-degree – are both titles, abbreviated as dr., so that it frequently leads to ambiguity. It could, therefore, be expected that doktor's vowel would not be prone to derounding.

[+]Age as a factor

A special case in support of the strong tendency to derounding is the pronunciation of the 2nd person singular pronoun u /y/ in Afrikaans. While it is hardly ever heard in casual Afrikaans, in formal instances it is still sometimes observable. In prayers, as is typically heard in sermons, u is frequently used. Interestingly, even in such extremely formal styles, derounding is not totally absent. This is particularly evident in the Afrikaans usage of young preachers. Wissing and Van Dijkhorst (2006) reports a RAP of 0.8 for derounding of u to [i] in the case of a 25-year old minister-in-training, in contrast to a RAP of <0.1 for a 65-year old.

Above-mentioned results regarding age as factor in the realm of derounding are supported by the findings of an unpublished study. In this study, 16 individuals divided into four age groups (20 – 39; 40 – 59; 60 – 79; 80+) and balanced as to gender, participated in a production test. Word pairs consisting of unrounded vowels and their rounded counterparts ( /i/ ~ /y/ (e.g. nies sneeze ~ nuus news); /ə/ ~ /œ/ (e.g. dis it's ~ dus thus), and /e/ ~ /ø/ (e.g. mees most ~ neus nose) were involved. Generally the oldest group of speakers succeeded best in distinguishing between especially /i/ and /y/, and between /ə/ and /œ/, followed by the groups 40 – 59, 60 – 79 and 20 – 39, in this order. Only /e/ ~ /ø/ rendered a somewhat mixed bag.

The results pertaining to the youngest group are underscored by another experiment, in which a group of twelve undergraduate students took part. Only three of them succeeded in making a clear distinction between /ə/ and /œ/; only two with respect to /i/ and /y/. Their performance regarding the third pair ( /e/ ~ /ø/) was somewhat better in that 50% of them were successful in this task.

[+]Gender as a factor

No conclusive experimental results are available in this regard. It might be reasonable, however, to expect female speakers, generally considered to be leaders in language, and, for that matter, sound change, to exhibit the process of vowel derounding more than males do.

[+]Ethnicity as a factor

Rather scant information is available on the presence or absence of derounding in varieties other than Standard Afrikaans (see Introduction). Wissing (1994), (1995), (2011), (2017), in general, points out that the neutralization of the contrast rounded ~ unrounded, in the case of the marked non-back vowels, is variable, despite claims by Links (1989) that the marked rounded front vowels are rounded without exception in the relevant varieties. Wissing (2011) shows that this is clearly not true. Van Wyk (1983) states that these vowels are totally unrounded in pidginised Afrikaans, namely in so-called Black Afrikaans (Afrikaans spoken by black South Africans). See also De Wet (1993) and Meiring and Retief (1991) for a similar view.

    printreport errorcite