Taalportaal - the digital language portal

Dutch
Frisian
Saterfrisian
Afrikaans

Show all

The resolution of hiatus between a monophthong and a following vowel

quickinfo

This topic deals with how hiatus between a monophthong and a following vowel is resolved. Between the vowels a glide is inserted, which ends up as the onset of the syllable headed by the following vowel.

Examples of the resolution of hiatus between a monophthong and a following vowel are given in (1):

Examples of the resolution of hiatus between a monophthong and a following vowel

Between a front vowel and a following vowel

hiaat

/hia:t/

[(hi.)(ja:t)]

gap

dyen

/di+ən/

[(di.)(jən)]

those ones

seeën

/se:+ən/

[(se:)(jən)]

seas

Between a central vowel and a following vowel

fluor

/flyɔr/

[(fly.)(jɔr)/(fly.)(wɔr)]

fluorine

duët

/dy+ɛt/

[(dy.)(jɛt)/(dy.)(wɛt)]

duet

ljuwe

/ljy+ə/

[(ljy.)(wə)]

folk, people

keuën

/kø:+ən/

[(kø:)(jən)]

(billiard) cues

Between a back vowel and a following vowel

silhûet

/-huɛt/

[-(hu.)(wɛt)]

silhouette

bliuwe

/bliu:+ə/

[(blju:)(wə)]

to stay

dowen

/do:+ən/

[(do:)(wən)]

pigeons

Hiatus is resolved here by the insertion of the glide [j] or [w], the choice between which is determined by the quality of the left-most vowel: [j] with a front and [w] with a back vowel. This means that the left-most vowel and the glide are homorganic.

The central glide [ɥ] therefore is expected to be inserted after the central vowels [y] and [ø]. Phonetically speaking this may be correct − though this has never been measured −, but due to the fact that [ɥ] is not an underlying segment of Frisian, the glide following /y/ and /ø/ is mostly interpreted as the familiar glide [j]Booij (1995:66) notes the same for Dutch).

/y/ is to be looked upon as a central vowel (see /y/ and /ø/ as centyral vowels), which means that in principle the glide following it can be interpreted as both the front glide [j] and the back glide [w]. This expectation is borne out by fluor /flyɔr/ fluorine and duët /dy+ɛt/ duet (in (1b)), which are pronounced as either [(fly.)(jɔr)] and [(dy.)(jɛt)] or as [(fly.)(wɔr)] and [(dy.)(wɛt)].

extra

The older loanword senuwachtich /se:nyw+axtəɣ/ nervous has the realizations senu [w]achtich and senu [j]achtich. Stem-final /w/ must have been reinterpreted as the inserted back glide [w], which paved the way for the alternate realization with the front glide [j] .

We then expect the same possibilities with respect to the vowel /ø/, but this is contrary to fact. The glide following /ø/ is always interpreted as [j], as in keuën /kø:+ən/ [(kø:)(jən)] [*(kø:)(wən)] (billiard) cues (see (1b) above). Does this mean that /ø/ is realized more as a front (hence: less rounded) vowel than /y/? Though this might be the case, as yet there is no independent phonetic evidence in support of it.

The examples hiaat /hia:t/ [(hi.)(ja:t)] gap, dyen /di+ən/ [(di.)(jən)] those ones, fluor /flyɔr/ [(fly.)(jɔr)/(fly.)(wɔr)] fluorine, duët /dy+ɛt/ [(dy.)(jɛt)/(dy.)(wɛt)] duet, ljuwe /ljy+ə/ [(ljy.)(wə)] folk, people, and silhûet /-huɛt/ [-(hu.)(wɛt)] silhouette in (1) show that a left-hand short vowel is lengthened in the context of vocalic hiatus. This is in line with the Rhyme Constraint, according to which the rhyme of a word-internal syllable occupies two structural phonological positions.

extra

There are no stems ending in the long half-open vowels /ɛ:/ and /ɔ:/ (these vowel do not occur in word-final position either, see word-final vowels). However, words like bêd /bɛ:d/ bed and pôt /pɔ:t/ pot; piss-pot, when realized with much emphasis, can be pronounced bisyllabically, as [(bɛ:)(jət)] and [(pɔ:)(wət)]. The insertion of the glide is in line with the cases in (1): [j] after the front vowel /ɛ:/ and [w] after the back vowel /ɔ:/.

extra

19th century words like folkspojesij (< poësij) popular poetry, pejesy (< poësij) poetry, and Matthéwes (< Mattheüs) Saint Matthew show that glide insertion used to be a less restricted process than it is nowadays, since the quality of the glide could also be determined by the right-hand vowel of the hiatus pair.

extra

The glide [j] shows up between the half close front vowel /e:/ and a following vowel. It causes raising of /e:/ to [i.], so that words like reëel /re:e:l/ real, actual, fideo /fide:o:/ video, ideaal /ide:a:l/ ideal, and Koreaan /ko:re:a:n/ Korean are realized as [ri.je:l], [fidi.jo:], [idi.ja:l], and [koəri.ja:n]. Compare the older loanword triater [tri.ja:tər] theatre, an adaptation of teater /te:a:tər/.

extra

The older loanwords in the table below make it clear that vowel reduction may render the context for glide insertion opaque (in the topmost five cases, u represents the vowel /y/):

Table 1

kejane (< gujano < guano) guano

mannewaar (< mannewaal < manuaal) manual

sitewaasje (< situaasje) situation

jannewaarje / jannejaarje (< januaarje) january

febrewaarje / febrejaarje (< februaarje) february

supplejant (< supplijant < suppliant) supplicant

aalwe (< alewee < aloëe) aloe

That the full vowel has reduced here is indicative of the fact that the glide is likely to be in the onset of the following syllable, for vowel reduction is strongly favoured in an open syllable (see vowel reduction and the properties of schwa).

The following loanwords have lost the triggering vowel for glide insertion altogether:

Table 2

offisjeel [ɔfisje:l] official

offisjeus [ɔfisjø:s] unofficial

spesjaal [spe:sja:l] special

These forms must have developed out of [ɔfisije:l], [ɔfisijø:s], and [spe:sija:l], cf the Dutch written forms officieel, officieus, and speciaal. The words in (2) carry stress on the final syllable, which must have caused pre-stress [i] to reduce first, followed by deletion. The resulting complex onset [sj] is fine. For spesjaal, for instance, all this must have proceeded as follows:

Table 3

/spe:si+aal/ → [(spe:)(si.)(ja:l)] → [(spe:)(sə)(ja:l)] → [(spe:)(sja:l)]

See Booij (1995:138-139) for the same phenomenon in Dutch, analyzed by him as deletion of the vowel /i/.

However likely this course of events may seem, not all words which at first sight are eligible for it did in fact undergo the same change. The words dieet /die:t/ diet and piano /pia:no:/ piano, for instance, are realized as [(di.)(je:t)] and [(pi.)(ja:)(no:)] piano, and not as [(dje:t)] and [(pja:)(no:)], though phonologically speaking there is nothing wrong with the latter. This might be a matter of Dutch influence, which is all the more likely, since loanwords enter Frisian through mediation of Dutch.

Loanwords, however, may also go their own way. This holds for pyjama /pia:ma/ pyjamas. The noun as such is pronounced as either [pija:ma], [pija:mə], [pəja:ma] or [pəja:mə]. But the diminutive pyjaamke − a word used by and to little children −, is realized as [(pja:m)(kə)].

References

Booij, Geert1995The phonology of DutchOxfordOxford University Press
Booij, Geert1995The phonology of DutchOxfordOxford University Press

⇑ Back to top

About this article

Author(s): Willem Visser

Category: Frisian Phonology

Keywords: hiatusmonophthongvowel

Version: 1.0, December 2015

Version history:

version	editor(s)	date	remarks
1.0	Willem Visser	December 2015

printreport errorcite

Jump to:

introduction
references
related articles