• Dutch
  • Frisian
  • Afrikaans
Show full table of contents

In the following section, /d/-deletion, the most well-known process of consonant reduction in Afrikaans, will be discussed.


/d/-deletion is a well-known and widely-occurring process in Afrikaans, characteristic especially of casual speech but far from being exclusive to this style. The faster a speaker speaks, the more probable the rule will be applied (the RAP will be relatively higher). As in most other cases, factors like the familiarity of the word in everyday speech and the frequency of usage are promoting factors. In connected speech, especially in unaccented sentence positions, this type of consonant reduction is typical. However, it is such a typical characteristic of Afrikaans that even in formal language it is present in abundance. Coetzee (Coetzee, A.E. 1991) provides an excellent overview of the literature on this topic, as well as a list of rule application probability indexes, based on a corpus of spoken Afrikaans. Other references are: (Le Roux, T.H.; Pienaar, P. de V. 1927), (De Villiers, M.; Ponelis, F.A. 1992), (Combrink, J.G.H.; De Stadler, L.G. 1987), (Wissing, D.P. 1982), (Wissing, D.; Van Dijkhorst, A. 2006), (Wissing, Daan 2017).

Of a number of consonant-reduction processes resulting in the restoration of the universally optimal CV-syllable structure pattern, /d/-deletion seems to be one of the most common ones in Afrikaans. See Consonant cluster simplification for an overview. As in the case with some other processes of consonant reduction, /d/-deletion is found in a variety of contexts, within lexical items as well as across word boundaries in connected speech. Examples of the former are function words like anderother, plural forms such as handehands and derived adjectives such as wildewild (from wild); also in compound words, especially those with a high frequency usage, such as the names of week days. In Donderdag (see Figure 1 below) /dɔndərdɑx/[dɔ:nərɑx]Thursday the second /d/ as well as the third /d/ y tend to be dropped quite frequently, even in more formal style, as is evident in the readings of news bulletins. Listen to the following example:

Figure 1: Oscillogram and spectrogram of Donderdag

[click image to enlarge]

Examples in connected speech mainly involve the article die when preceeded by mainly function words ending on a sonorant consonant e.g. in a phrase like van diefor the .

In the following sections, cases of /d/-deletion like those above are provided and discussed in detail. As far as possible, examples have been extracted from the datasets mentioned in the Introduction to phonological processes, mostly from the RSG database.

[+] /d/-deletion in monomorphemes

This type of consonant reduction takes place primarily in multisyllabic words where /d/ is preceeded by a sonorant consonant and followed by a vowel, mainly schwa /ə/. A first approximation of the /d/-deletion rule can, therefore, be formalised loosely and in over-generalised form as Rule (1):

(1) /d/ ➜ Ø / {... V' + {/n l r/} $ _ /ə/ ...}

Note that Ø = zero; V' = an accented vowel; $ = syllable boundary; "..." indicates the presence or absence of other segments; _ indicates the position of /d/ in the form prior to deletion taking place.

As mentioned above, in reality this rule is an over-generalisation: "Vowel" in fact is limited to a subset of mainly non-high vowels and to practically no diphthongs.

In the following, examples and discussion of /d/-deletion will be largely restricted to cases where the sonorant consonant /n/ is involved; /r/ and /l/ will only occasionally be mentioned.

The role of frequency of usage seems clear when comparing the RAP results of the following three cases :

  1. a[nd]era[n]erother
  2. so[nd]erso[n]erwithout
  3. ha[nd]elha[n]eltrade

Coetzee (1991) mentions an overall average RAP 0.65 in the case of -der (as in 1 - 3 above), but while 1 and 2 are instances characterised by an extremely high RAP, even in formal speech, 3 is not. This illustrates the fact that phonetic content is not the only deciding factor determining the degree of rule application. All these words are well-known in Afrikaans, but ander and, to a lesser extent, sonder, are unlike handel, characterised by a very high frequency of usage. This fact is evident in other cases where /d/ gets deleted, to be discussed in the following sections. . In the above three examples, as in many other cases, /d/-deletion results, via resyllabification, in the preferred syllable pattern (C)V$CV, from (C)VC$CV - see Introduction to phonological processes for more on this.

[+] /d/-deletion in plural forms

/d/ in plural forms ending on -e, which have a high frequency of usage, and in relation to which the underived form ends on /d/ - said /d/ is regularly deleted, rendering the same syllable pattern as in the case of monomorphemes, viz. (C)V$CV. Examples are han[d]+ehands and hon[d]+edogs, again with retention of the preferred syllable pattern. Usually the vowel preceeding /n l r/ is short; words like aardeearth and woordewords, however, contain long vowels; in eindeend it is a diphthong that precedes the sonorant consonant. In these three last words, /d/ is readily deleted in casual speech although seldom in formal style.

[+] /d/-deletion in adjectival derivations

Similar to the preceeding cases, /d/ is often deleted in derived adjectives in attributive position, e.g. ron[d]+eround and wil[d]+ewild wild.

A prototypical example of /d/-deletion in derived adjectives is the following: in the RSG news bulletin dataset (Introduction to phonological processes) none of the 33 occurrences of volgendenext were read with retention of /d/. In casual speech, the same tendency is widely observable. The adverb gedurendeduring behaves similarly. On the other hand, a well-known and reasonably frequent word like bekendewell-known is seldom heard without [d].

In some commonly used adjectives containing morphemes like -elik, -eling, -erig e.g. in, respectively vriendelikfriendly, mondelingoral and ronderigroundish, /d/ tends to be deleted.

The lack of a high frequency of usage is, undoubtedly, the reason for the blocking of this rule in words like akkommoderendeaccommodating and innoverendeinnovating.

In is notable that /d/ does not undergo deletion in other phonetic contexts e.g. when /d/ is preceeded by obstruents such as /f s/. Therefore adjectives like gekerfdecarved or afgesponsdesponged down are not effected: *gekerfe, *afgesponse.

[+] /d/-deletion in ordinals

In casual speech, /d/-deletion takes place in ordinals ending on -de, in particular in derde, vierde, sewende, negendethird, fourth, seventh, ninth, all words of a relatively high frequency of usage. In formal speech (e.g. in the RSG dataset) this is not the case.

[+] Deletion of /d/ in function words

/d/ as onset of the function word diethe tends to be deleted in the same envionment as indicated above. This regularly happens in casual as well as formal connected speech. Less commonly, /d/ also undergoes deletion in datthat, in some cases even in complex words such as voordatbefore, even in omdatbecause, despite /m/ not being part of the primary set of sonorant consonants triggering /d/-deletion and as stipulated above.

Some prominent examples, found in the RSG dataset, of /d/-loss in die, are listed below . In order to give a picture of the extensive spread of this phenomenon in Afrikaans, some statistics are mentioned.

In total, 786 instances of the phrase in diein the were present. A subset of these - from only one reader - contains 21 cases. Of these, all but two in die phrases underwent deletion of /d/; die in both the exceptions was emphasised i.e. written as dié, denoting this rather than the. Other highly frequently used phrases of this kind include the following, all including a function word and all showing high RAP indexes:

/n/ + /d/: aan diefor the, en dieand the, kan diecan the, kon diecould the.

/l/ + /d/: al dieall the, sal dieshall the, wil diewill the.

/r/ + /d/: agter diebehind the, maar diebut the, oor dieover the, vir diefor the, voor diein font of the.

Coetzee (1991) reports RAP's of 0.52, 0.29 and 0.42 for these respective sub-classes for a wider range of examples.

Finally, a category of commonly used verbs having /n/ as coda plus die following, shows similar behavior: doen diedo the, sien diesee the, speel dieplay the, voel diefeel the, hoor diehear the, probeer dietry the.

The adverb ditit behaves in a similar way in connected speech to die, albeit in a much more restrictive manner; compare: en ditand it, kan ditcan it, kon ditcould it, sal ditshall it, wil ditwill it, maar ditbut it, verstaan ditunderstand it, voel ditfeel it, hoor dithear it.

[+] Cases other than /n l r/

/d/ is frequently deleted in the function word diethe when in combination with commonly used words which have an obstruent as coda. This happens predominantly in casual speech. A few representative examples are:

op dieon the, of dieif the, mag diemay the, as diewhen the.

Coetzee's (1991) RAP's for obstruents followed by [d] are: [p] = 0.77, [k] = 0.55, [s] = 0.65, [f] = 0.23.

  • Coetzee, A.E1991'n Nuwe kyk na [d]-weglating in Afrikaans.Tydskrif vir Geesteswetenskappe31113-127
  • Combrink, J.G.H. & De Stadler, L.G1987Afrikaanse fonologie.Southern Boekuitgewers
  • De Villiers, M. & Ponelis, F.A1992Afrikaanse klankleer.Tafelberg
  • Le Roux, T.H. & Pienaar, P. de V1927Afrikaanse fonetiek.Juta
  • Wissing, D. & Van Dijkhorst, A2006Is spreekstyl en spreektempo sinonieme?: 'n fonologiese ondersoek.Southern African Linguistics and Applied Language Studies24217-233
  • Wissing, D.P1982Algemene en Afrikaanse generatiewe fonologie.Macmillan
  • Wissing, Daan2017FonologieVan Schaik
Suggestions for further reading ▼
  • Dutch
  • Frisian
  • Afrikaans
  • Rhotacism
    [91%] Afrikaans > Phonology > Phonological Processes > Consonant related processes
  • Nasal assimilation
    [89%] Afrikaans > Phonology > Phonological Processes > Consonant related processes
  • r-deletion
    [89%] Afrikaans > Phonology > Phonological Processes > Consonant related processes
  • Quality alternation of back vowels
    [88%] Afrikaans > Phonology > Phonological Processes > Vowel related processes
  • Nasalization
    [88%] Afrikaans > Phonology > Phonological Processes > Vowel related processes
Show more ▼
  • Dutch
  • Frisian
  • Afrikaans
  • In prenominal position
    [87%] Frisian > Morphology > Inflection > Adjectives
  • Cardinal numbers
    [87%] Frisian > Morphology > Inflection > Numerals
  • -mannich
    [86%] Frisian > Morphology > Word formation > Derivation > Suffixation > Nominal suffixes > Noun as base
  • Weak verbs
    [86%] Frisian > Morphology > Inflection > Verbs
  • -DIM (diminutive)
    [86%] Frisian > Morphology > Word formation > Derivation > Suffixation > Nominal suffixes > Noun as base
  • Prefixation
    [77%] Afrikaans > Morphology > Word formation > Affixation
  • Circumfixation
    [77%] Afrikaans > Morphology > Word formation > Affixation
  • Non-native affixes
    [77%] Afrikaans > Morphology > Word formation > Affixation
  • Affixation
    [76%] Afrikaans > Morphology > Word formation
  • Conversion
    [76%] Afrikaans > Morphology > Word formation
Show more ▼
  • Dutch
  • Frisian
  • Afrikaans
  • Appendix: alternative means of expressing spatial relations
    [87%] Dutch > Syntax > Adpositions and adpositional phrases > 1 Characteristics and classification > 1.3. A semantic classification of adpositional phrases > 1.3.1. Spatial adpositions
  • 5.4. N-ellipsis
    [86%] Dutch > Syntax > Adjectives and Adjective Phrases > 5 Attributive use of the adjective phrase
  • 1.3. Inflection
    [86%] Dutch > Syntax > Verbs and Verb Phrases > 1 Characterization and classification
  • 3.3.2. Accusative/PP alternations
    [86%] Dutch > Syntax > Verbs and Verb Phrases > 3 Projection of verb phrases II:Verb frame alternations > 3.3. Alternations of noun phrases and PPs
  • Special cases
    [85%] Dutch > Syntax > Nouns and Noun Phrases > 5 Determiners: articles and pronouns > 5.2. Pronouns > 5.2.3. Demonstrative pronouns
  • Attribution
    [90%] Afrikaans > Syntax > Introduction to Adjective Phrases
  • Constructions with APs
    [88%] Afrikaans > Syntax > Introduction to Adjective Phrases > Characteristics and classification
  • Mood
    [87%] Afrikaans > Syntax > Introduction to Verb Phrases > Characterization and classification
  • Root modality
    [87%] Afrikaans > Syntax > Introduction to Verb Phrases > Characterization and classification > Modality
  • Finite declarative complement clauses: construction forms
    [86%] Afrikaans > Syntax > Introduction to Verb Phrases > Complement clauses > Finite declarative complement clauses
Show more ▼
This is a beta version.