• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
Regressive Voice Assimilation: type 1

Regressive voice assimilation (RVA) is a more heterogeneous phenomenon than regressive place assimilation. There are two types, each with a different class of inductive and target consonants. The first type of RVA, which is the subject of this topic, entails that a plosive adopts the voice specification of a following voiced plosive. An illustrative example is the separable verb útgean /yt#ɡɪən/ to go out, which is realized as [ydɡɪən], so with the voiced sequence [-dɡ-].


Examples of the first type of regressive voice assimilation are given in the table below:

Examples of assimilation of a plosive to a following voiced plosive

Table 1
útgean [dɡ] to go out
in hut bouwe [db] to build a cabin
opdwaan [bd] to open up; to turn on, to switch on
(dat is) knap gien [bɡ] (all that went) pretty well
drekboel [ɡb] muddy, sludgy mess
(dat ha 'k) faak dien [ɡd] (I've) often done (that)

A general constraint on obstruent sequences is that the obstruents must agree in voicing (see onset: sequences of two obstruents), a pattern which this kind of RVA adheres to.


The fricative /v/ is often strengthened in word-initial position (see the labial fricatives). This may proceed so far that it can act as a voiced plosive. In some dialects then /v/ triggers this kind of RVA, which is exemplified in the table below:

Table 1
wiet waar /viət va:r/ [vi.əd va:r] wet, rainy weather
opwine /op#vinə/ [obvinə] to wind (up); to excite
do koest wol gek wêze /ɡɛk vɛ:zə/ [ɡɛɡ vɛ:zə] have you gone out of your mind? (lit. it looks as if you have gone crazy)

A voiceless plosive seems to be able to influence the voice specification of a plosive to its left, even though both plosives are not adjacent. Examples, nearly all of them with /b/, are given below:

Table 2
brutaal ~ bertaal pertaal impudent, impertinent
beton peton concrete
bottines ~ betines petines high-lows
sainte boutique ~ santebetyk santepetyk the lot, the caboodle
abrikoas ~ abrekoas apprekoas apricot
agremint akkremint garnishing
brouette ~ brewet prewet type of wheelbarrow
by túltsjes ~ betúltsjes petúltsjes in jest, jokingly
idere kear ~ idekear itekear each time, every time
The forms in the second column derive from those in the first column. If there is regressive assimilation here at all, it is quite different from the kind of assimilation which is the subject of this topic: a) the change is not from voiceless to voiced, but he other way around, b) both plosives are not adjacent, but they are separated by a vowel or the liquid /r/. A condition seems to be that they must not be separated by an obstruent (a fricative or a plosive). The plosives are separed by the vowel schwa, which entails that the voiceless plosives end up as the onset of un unstressed syllable. It also entails that the original full vowel must have undergone reduction before the devoicing of the plosive was possible. In all likelihood, this kind of obstruent adaptation is not to be viewed as an instance of regressive voice assimilation, but as one of consonant harmony.

    printreport errorcite