• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
/d/-insertion in the sequences /nər/, /lər/, and /rər/
quickinfo

The consonant /d/ is inserted between one of the consonants /n/, /l/ or /r/ and /ər/. The commonest contexts in which the sequences /nər/, /lər/, and /rər/ appear are the comparative of adjectives, subject names derived from verbs, and inhabitant names and relational adjectives derived from place names. Examples of /d/-insertion in comparatives are tin - tinder thin - thinner, fûl - fûlder intense - more intense; sharp - sharper, and fier - fierder far - farther, further. There is, however, an asymmetry here. Following /n/ and /l/, insertion is optional (and receding in extension), whereas it is obligatory (and fully productive) following /r/. This links up with a general ban on sequences in which /r/ flanks a short vowel on either side.

readmore

When preceding the sequence /ər/, the consonant /d/ is inserted, provided it is itself preceded by one of the consonants /n/, /l/ or /r/ (see Sytstra (1856:51-52), Colmjon (1863:51), Van Blom (1889:91), Sipma (1913:62), Sytstra and Hof (1925:49), Sipma (1949:20), Fokkema (1967:46)), Tiersma (1979:125-129), Tiersma (1980:12-14), Tiersma (1985:46), Tiersma (1999:31-32), Visser (1997:161-163), Hoekstra (2001:87), Popkema (2006:78)). The commonest contexts in which the above sequences appear are the comparative of adjectives, subject names derived from verbs, and inhabitant names and relational adjectives derived from place names. These are exemplified in (1):

1
Examples of the insertion of /d/ between /{n,l,r}/ and /ər/
a. With stems ending in /n/
a.1 In the comparative of adjectives
brún ~ brúnder brown ~ browner
grien ~ griender green ~ greener
tin ~ tinder thin ~ thinner
skruten ~ skrutender timid ~ more timid
tefreden ~ tefredender satisfied ~ more satisfied
a.2 In subject names
post#rinder postman (cf. rinn(e) walk, go )
priis#winder prizewinner (cf. winn(e) win )
iel#slynder (great crested) grebe (cf. slin(e) eat, feed, stuff )
a. 3 In inhabitant names and relational adjectives
Eastereinder inhabitant of, related to Easterein
Westereinder inhabitant of, related to De Westerein
Peinder inhabitant of, related to De Pein
b. With stems ending in /l/
b.1 In the comparative of adjectives
fol ~ folder full ~ fuller
fûl ~ fûlder intense ~ more intense; sharp ~ sharper
giel ~ gielder yellow ~ yellower
keal ~ kealder bald ~ balder; bare ~ barer
koel ~ koelder cool ~ cooler
mâl ~ mâlder silly ~ sillier; mad ~ madder
smel ~ smelder narrow ~ narrower
stil ~ stilder quiet ~ quieter; silent ~ more silent
krigel ~ krigelder energetic ~ more energetic
b.2 In subject names
dwers#kúlder fisherman who 'dwerskulet' (cf. dwerskúl(je) fish with a 'dwerskule' )
gear#stalder compiler (cf. gearstall(e) to compile )
lulder one who talks bullshit (cf. lull(e) talk bullshit )
piano#spylder piano player (cf. spyl(je) to play )
pielder someone who occupies himself with light, small or lowly-qualified activities (cf. piel(e) occupy oneself (in a relaxed way and without compelling reasons) )
angelder angler (cf. angel(je) to angle )
dobbelder dicer (cf. dobbel(je) to dice )
gûchelder conjurer, magician (cf. gûchel(je) to conjure, do tricks )
skeakelder switch (cf. skeakel(je) to link; to connect; change gear )
sutelder street-trader, hawker (cf. sutel(je) to hawk )
b.3 In inhabitant names and relational adjectives
Fochtelder inhabitant of, related to De Fochtel
Oeriselder inhabitant of, related to the province of Oerisel
Sondelder inhabitant of, related to Sondel
Terwispelder inhabitant of, related to Terwispel
Twizelder inhabitant of, related to Twizel
Skarlder inhabitant of, related to Skarl
[Note: Though the spelling suggests otherwise, Skarl ends in /-rəl/ in underlying representation (see word-final sequences of a liquid and a liquid).]
c. With stems ending in /r/
c.1 In the comparative of adjectives
bizar ~ bizarder bizarre ~ more bizarre
djoer ~ djoerder expensive ~ more expensive
fier ~ fierder far ~ farther, further
klear ~ klearder clear ~ clearer
raar ~ raarder odd, strange ~ odder, stranger
toar ~ toarder barren ~ more barren; withered ~ more withered
nuver ~ nuverder unusual ~ more unusual
suver ~ suverder pure ~ purer
c.2 In subject names
bearder one who makes a fuss (cf. bear(e) make a fuss )
behearder manager (cf. behear(e) to manage )
bestjoerder driver; director (cf. bestjoer(e) to drive; to run )
feest#fierder partygoer (cf. feestfier(e) to celebrate )
hierder renter; tenant (cf. hier(e) to rent; to hire )
hús#warder caretaker (cf. húswar(je) look after the house )
learder studious person (cf. lear(e) to study, to learn )
oan#fierder leader (cf. oanfier(e) to lead )
ôf#stjoerder sender (cf. ôfstjoer(e) send away )
skieppe#skearder sheepshearer (cf. skieppeskear(e) shear sheep )
slange#beswarder snake-charmer (cf. beswarr(e) to conjure )
ta#hearder listener (cf. tahearr(e) to listen (to) )
tsjinst#wegerder conscientious objector (cf. weger(je) to refuse )
walfisk#farder whaler (cf. farr(e) to sail, to navigate )
heisterder one who rushes around (cf. heister(je) rush around )
kankerder grouser, grumbler (cf. kanker(je) to grouse, to grumble )
plonderder plunderer, looter (cf. plonder(je) to plunder, to loot )
sangerder bore, nag (cf. sanger(je) to nag, to wine )
c.3 In inhabitant names and relational adjectives
Eksmoarder inhabitant of, related to Eksmoarre
Gaastmarder inhabitant of, related to De Gaastmar
Langwarder inhabitant of, related to Langwar
Legemarder inhabitant of, related to Legemar
Molkwarder inhabitant of, related to Molkwar
Sumarder inhabitant of, related to Sumar
Waskemarder inhabitant of, related to Waskemar

extra

The phrasing above that /d/ is inserted when preceding the sequence ər has been chosen deliberately. It is noted in the literature − see Booij (1995:74) for the same phenomenon in Dutch − that /d/ is also inserted between a stem ending in /n/, /l/, or /r/ and the suffixes -erich, -erij, and -ernôch, as in stinderich wavery (from stinn(e) to waver), pielderich dilettanteish, amateurish (from piel(e) occupy oneself (in a relaxed way and without compelling reasons)), tsierderich quarrelsome (from tsier(e) to quarrel), strúnderij rummaging about (from strun(e) rummage about), seurderij nagging, complaining (from seur(e) to nag, to complain), djoerdernôch quite expensive; in fact too expensive (from djoer expensive), kleardernôch clearly, unequivocally (from klear clear). As with -er, insertion is optional with a stem ending in /n/ or /l/ and obligatory with an /r/-final one (see also below).

extra

The inhabitant name and the relational adjective of the place name Frjentsjer are not Frjentsjerder, as is to be expected, but Frjentsjerter. The name Frjentsjer has developed from franaeker lord's acre, so its final /r/ is etymologically correct. More often than not, however, the name is pronounced with a final /t/, in front of which /r/ has deleted. This /t/-final form then must underly the inhabitant name and the relational adjective.

The above, however, is in need of qualification: /d/-insertion is optional with the sequences /nər/ and /lər/, but obligatory with /rər/. The derived forms in (1a,b) therefore also occur without /d/. As a matter of fact, younger generations realize ever fewer forms with /d/, to the extent that /d/-insertion has virtually stopped applying.

extra

Place names ending in /{n,l,r}/ all have inhabitant names and relational adjectives ending in -der. As they form a set which hardly increases, if at all, it cannot be tested whether /d/-insertion still or no longer applies in the context of place names.

This does not mean, however, that the sequences /-ndər/ and /-ldər/ do not occur any more, for some words retain a once-inserted /d/, examples of which are provided in (2) (see also Tiersma (1980:13-14)):

2
Examples of non-alternating words in which inserted /d/ is retained
daalder one guilder and a half (cf. English dollar , German Taler )
kelder cellar (cf. English cellar , German Keller )
pylder pillar (cf. English pillar , Dutch pilaar )

These are simplex words, which do not alternate with /d/-less forms, so /d/ has been reinterpreted as part of the underlying representation.

Such an explanation, however, is not available for all words with /d/. As a first example, take bjinder and bjinner, both deriving from the verb bjinn(e) to scrub. That bjinder is still in use is likely to be due to the fact that it has developed a meaning different from that of bjinner. Whereas the former has become an instrument noun, meaning scrubbing-brush, scrubber, the latter is, and remains, a pure agent noun, meaning one who scrubs. Meaning differentiation thus may be a source of preservation. In Dutch the same holds for the relation between boender scrubbing-brush, scrubber and boener one who scrubs− deriving from the verb boen(en) to scrub−, which may have had an influence on the Frisian word pair bjinder - bjinner.

Take, as a second example, the word minder less, fewer. This is the comparative of the quantifier min little, few. The latter, however, can only be used in combination with the adverbs te too and sa so (te min too little, too few, sa min mooglik as little, as few as possible). In other contexts, a small quantity of something is indicated by phrases like net folle not much, many, in pear a few (literally: a pair), in bytsje a little bit, and the like. This means that minder does not have a transparent compositional meaning, viz. min plus comparative, which in all likelihood is the reason why it is retained as such. Again, the same may hold for Dutch minder less, fewer, the comparative of weinig little, few.

As a third example, take compounds ending in -gunder one who grants another one something (derived from the verb gunn(e) to grant): goegunder donor, supporter, kweagunder envious, jealous person, and nearinggunder one who gives you his custom.

Finally, take tsjoender magician, sorcerer, wizard. Though it seems to have a transparent relation with the verb tsjoen(e) work magic, its /d/ is retained; this also goes for the derived noun tsjoenderij magic, sorcery, witchcraft. An explanation for /d/'s retention here is not readily available.

extra

Next to less, fewer, the word minder also means worse, in which case it is the comparative of min bad. Despite the fact that minder has a transparent compositional meaning here, its /d/ is obligatory, for the form *minner does not occur.

extra

The word sigeuner gypsy has the variant sigeunder; word-final -er must have been reinterpreted as a suffix here.

extra

Frisian has also borrowed words, with /d/ and all, from Dutch. Two examples are einder horizon and stander stand (of a bicycle, for instance).

extra

The prefix aller- is attached to the superlative of adjectives, either to denote a still higher degree than the superlative does or to denote a high degree just like that. In the great majority of instances, however, the variant alder- is used, as in (it) alder#moaist most beautiful by far and alder#aardichst very, very nice. In fact, according to Sytstra and Hof (1925:49)alder- is the only form possible. It is also used in case of prefix doubling, as with it wie alder-alderferskuorrendste min waar the weather was horribly bad (also in nominal use, like do bist myn alderalderste you are the one I love most of all). Confer the variant alderbalder (< alder by alder), as in alderbaldermoaist very, very beautiful.

extra

alder- is also the first part of the adjective alderhande, which is used in denoting a group of comparable objects, as in alderhande stuollen chairs of all sorts/kinds, all sorts/kinds of chairs, and alderlei, which by and large has the same function, as in alderlei drokte all sorts/kinds of fuss, bother. It is also found in Alderheljen All Saints'Day and Aldersielen All Soul's Day. This alder- is related to the numeral al/alle all.

Additional evidence for the difference between the sequences /nər/ and /lər/ on the one and /rər/ on the other hand, is provided by some cliticization facts. If the third person singular preterite form of a strong/irregular verb ending in /-{n/l}/ is followed by the clitic form of the third person singular masculine personal pronoun, both allomorphs of the latter, i.e. er /ər/ and der /dər/, can show up, whereas only der is allowed following an /r/-final stem (Visser (1997:162-163)). See the examples in (3):

3
Examples of the occurrence of the clitics 'er' and 'der' following a preterite stem ending in {/n,l,r/}
a. With stems ending in /n/
dat fûn er/der dêr net that found he there not he didn't find that there
b. With stems ending in /l/
dêr foel er/der dea del there fell he dead down there he dropped dead
c. With stems ending in /r/
sa fear der/*er fuort so sailed he away that's how he sailed away
sok iten sear der/*er altyd such food cooked he always that's the kind of food he always cooked

Since er and der must be considered as independent variants (see cliticization), there is no /d/-insertion here. The above pattern, however, is indicative of the difference between the sequences /{n/l}ər/ and /rər/.

What is behind the insertion of /d/ between /{n/l}/ and /ər/ or, put differently, what is it that makes the sequences /nər/ and /lər/ eligible for /d/-insertion? Note first that /{n/l/r/}/, being sonorant consonants, share many features and second that they are separated by schwa here. Since the latter is an essentially featureless, minimally specified vowel (see schwa's phonological representation), the consonants in these sequences are virtually adjacent. Tiersma (1980:12) therefore considers /d/-insertion a process which is functionally equivalent to a dissimilation process in that it prevents sequences of (sonorant) consonants which are identical or very similar to one another. If the sequences /nər/ and /lər/ result from morphological processes, they are repaired by inserting the obstruent which shares the largest number of features with the sonorant consonants, namely d. The dissimilatory effect then seems to reside in the fact that /d/ is an obstruent, in strong contrast with the sonorant consonants by which it is surrounded. Sonorancy, however, cannot be the only factor here, for the sequences /mər/ and /ŋər/ are fine as such, i.e. they are not broken up by /d/ or any other obstruent.

Now, /{n,l,r}/ share more than a high degree of sonorancy. They all are coronal consonants, which suggests that their place specification might (also) play a role. Is /d/-insertion triggered then by the fact that the consonants of the sequences /nər/ and /lər/ have the same place of articulation? Note that this would explain the non-occurrence of /d/-insertion with respect to the sequences /mər/ and /ər/, which consist of the bilabial/velar nasal consonants /{m/ŋ}/ and the coronal liquid (rhotic) /r/. The same place specification of the segments in the sequences /nər/ and /lər/ causes an OCP violation, be it only a slight one, since after all the consonants are separated by schwa. However, since /d/ is a coronal consonant as well, nothing seems to be gained by its insertion. At the level of sonorancy, however, /d/-insertion does make for a better contour within these sequences, for the value -sonorant of /d/ contrasts with the +sonorant value of the surrounding consonants. Be that as it may, it seems highly unlikely that an OCP violation with respect to the place specification of adjacent consonants can be repaired by establishing a contour at the level of sonorancy.

As noted above, the younger generations realize ever fewer forms ending in /-ndər/ and /-ldər/, to the extent that /d/-insertion has virtually stopped applying. This may have to do with the status of the process, that is, its unclear constraints-and-repair-status (see the previous paragraph) may have furthered its disappearance from the grammar.

With respect to the sequence /rər/, /d/-insertion is obligatory, a fact which cannot but yield a different explanation than for the sequences /nər/ and /lər/. This will be discussed now (see also Visser (1997:161-163)). The liquid /r/ cannot flank a short monophthong on either side (see Booij (1995:43) as to Dutch, whereas it is maintained in Booij (1995:73) that this is a cross-linguistic tendency). This constraint on the occurrence of /r/ might be formalized as in (4):

/r/-constraint
* (rVr)σ

The constraint expresses the fact that raar /ra:r/ odd, strange, queer, rear(e) /rɪər/ stir, rier /riər/ heifer, skroar /skroər/ tailor, clothier, roer /ruər/ rudder, treur(je) /trøər/ sorrow, mourn are possible, whereas *rar /rar/, *rir /rɪr/, *ryr /rir/, etc. are not. Words ending in /-rər/ are excluded as well (be they simplex or derived).

extra

The interjection (wild oath) harrekrarre /harəkrarə/ ugh! is an exception.

extra

The fact that /r/ can flank (short) rising diphthongs, as in beruor(je) /bərwor/ arrange, fix (up) and (adapted) loan words ending in -ear(je) ( /-jɛr/), like orear(je) /o:rjɛr/ orate, declaim and kastrearje /kastrjɛr/ castrate, is an indication that a rising diphthong consists of two components at the segmental level.

The /r/-constraint has repercussions for the realization of complex words and combinations of host word + clitic, and for the operation of phonological rules.

Firstly, the constraint implies that /d/-insertion is obligatory if the stem preceding /ər/ ends in /r/. This is exemplified in (1a) of /r/-deletion in complex words derived with a suffix which contains the vowel schwa and in (1c) above (see Booij (1995:73) for examples of the same phenomenon in Dutch). This implies that

  1. the agent noun of hier(e) /hiər/ hire, rent, lease is hierder tenant, renter, hirer (not *hierer);
  2. the comparative of djoer /djuər/ expensive, costly is djoerder (not *djoerer);
  3. only the adjective tsierderich quarrelsome and the noun tsierderij arguing, quarrelling can stand next to the verb tsier(e) /tsiər/ quarrel (and not *tsiererich and *tsiererij);
  4. only the inhabitant names/relational adjectives Langwarder and Sumarder can stand next to Langwar and Sumar (and not *Langwarrer and *Sumarrer).

The same holds for the cliticization case in (3c) above, repeated here as (4): if the preterite stem ends in /-r/, only the clitic allomorph der is possible:

4
Examples of the insertion of /d/ in the context of cliticization
sa fear der/*er fuort
sok iten sear der/*er altyd

Secondly, both the insertion of /d/ between stem-final /r/ and /ər/ and the subsequent deletion of /r/ before inserted /d/ are obligatory processes, examples of which are also given in /r/-deletion in complex words derived with a suffix which contains the vowel schwa. This means that only the sequence /-dər/ remains, see (5):

5
The realization of forms ending in the sequence /-rdər/
hierder [hiədr̩]
djoerder [djuədr̩]
tsierderich [tsiədərəx]
tsierderij [tsiədərɛj]
fearder [fɪədr̩]
searder [sɪədr̩]
Langwarder [laŋvadr̩]
Sumarder [səmadr̩]

Since /d/ is inserted before more than one suffix and in phonological words consisting of a host word and a clitic, the process can be described in purely phonological terms.

extra

As noted, /d/ is obligatorily inserted between /r/ and -er and the inserted /d/, in its turn, causes the obligatory deletion of /r/. This means that in forms like those in (5) /r/ is recoverable via /d/, a segment belonging to neither stem nor suffix.

extra

When following a stressed and preceding an unstressed vowel, /d/ can (could) turn into /r/, which gives (gave) rise to a great deal of dialectal variation (see intervocalic /d/ and /r/). In some dialects then the verbs wurd(e) /vød/ become and hâld(e) /hɔ:d/ to hold, to keep became wurr(e) /vør/ and hâr(e) /hɔ:r/. But although the verbs biede /biəd/ to offer and siede /siəd/ to boil; to cook turned into bier(e) /biər/ and sier(e) /siər/ dialectally, the agent nouns bieder /biəd+ər/ bidder and (iten)sieder /(itən#)-siəd+ər/ who does the cooking could not possibly become *bierer and *(iten)-sierer. The impossibility of /d/-rhotacism here links up with the ill-formedness of the sequence /*rər/.

An equally revealing case is the noun ljedder /ljɛdər/ ladder; scale, which has the dialectal variant ljedde /ljɛdə/. But whereas the form *ljerrer (from ljedder) does not occur, the form ljerre (from ljedde) does.

extra

The ending -earje of loanverbs can be realized as [ɪərjə] or [jɛrjə], according to dialect, so the pronunciation of sollisitearje apply (for) is either [sɔlisitɪərjə] or [sɔlisitjɛrjə]. In the great majority of cases, the agent nouns accompanying these verbs are realized with [ɪə] and not with [jɛ], which means that sortearder sorter− from sortearje to sort (out)− is predominantly pronounced as [sɔrtɪədr̩] (and not as [sɔrtjɛdr̩]). This preference for [ɪə] may be ascribed to the fact that /r/-deletion, which is obligatory here, fares better when following a long vocalic sequence than when following a short vowel (see /r/-deletion).

References
  • Blom, Philippus van1889Beknopte friesche Spraakkunst voor den tegenwoordigen tijdLeeuwardenJ.W. Muller
  • Booij, Geert1995The phonology of DutchOxfordOxford University Press
  • Booij, Geert1995The phonology of DutchOxfordOxford University Press
  • Booij, Geert1995The phonology of DutchOxfordOxford University Press
  • Booij, Geert1995The phonology of DutchOxfordOxford University Press
  • Colmjon, Gerben1863Beknopte Friesche spraakkunst voor den tegenwoordigen tijdZijlstra, Joure
  • Fokkema, Klaas1967Beknopte Friese SpraakkunstFryske Akademy, Leeuwarden
  • Hoekstra, Jarich2001Standard West FrisianMunske, Horst Haider, Århammar, Nils, Hoekstra, J.F., Vries, O., Walker, A.G.H., Wilts, O. & Faltings, V.F. (eds.)Handbuch des Friesischen / Handbook of Frisian StudiesMax Niemeyer83-98
  • Popkema, Jan2006Grammatica FriesUtrecht/ LjouwertUitgeverij Het Spectrum BV Prisma Woordenboeken en Taaluitgaven/ Fryske Akademy
  • Sipma, Pieter1913Phonology and Grammar of Modern West FrisianLondon, New YorkOxford University Press
  • Sipma, Pieter1949Ta it Frysk IIIR. van der Velde, Ljouwert
  • Sytstra, Harmen S1856Klank- en Schriftleer der Friesche taal, of eerste afdeeling der Friesche SpraakkunstLeeuwarden
  • Sytstra, Onno H. & Hof, Jan J1925Nieuwe Friesche SpraakkunstLeeuwardenR. van der Velde
  • Sytstra, Onno H. & Hof, Jan J1925Nieuwe Friesche SpraakkunstLeeuwardenR. van der Velde
  • Tiersma, Pieter M1979Aspects of the phonology of Frisian based on the language of GrouMeidielingen fan de stúdzjerjochting Frysk oan de Frije Universiteit yn Amsterdam4
  • Tiersma, Pieter M1980The lexicon in phonological theoryBloomington, Ind.Indiana University Linguistics ClubThesis
  • Tiersma, Pieter M1980The lexicon in phonological theoryBloomington, Ind.Indiana University Linguistics ClubThesis
  • Tiersma, Pieter M1980The lexicon in phonological theoryBloomington, Ind.Indiana University Linguistics ClubThesis
  • Tiersma, Pieter M1985Frisian reference grammarDordrechtForis Publications
  • Tiersma, Pieter M1999Frisian Reference GrammarAfûk, Ljouwert
  • Visser, Willem1997The Syllable in FrisianVrije Universiteit AmsterdamThesis
  • Visser, Willem1997The Syllable in FrisianVrije Universiteit AmsterdamThesis
  • Visser, Willem1997The Syllable in FrisianVrije Universiteit AmsterdamThesis
printreport errorcite