• Dutch
  • Frisian
  • Afrikaans
Show full table of contents
The realization of /i:/ and /u:/

This section deals with the long close vowels /i:/ and /u:/, which are evermore being realized as the centring diphthongs [iə] and [uə], especially when preceding a coronal consonant.


Sipma (1913:9, §4) observes that the close long vowels /i:/ and /u:/ can be realized as [i.j] and [u.w] (he does not mention /y:/, which has a low type frequency). According to Riemersma (1979:19-20), this realization only occurs in case /i:/ and /u:/ precede /s/ or /t/, whereas Boersma and Van der Woude (1972:48-49) consider the long close vowels as homogeneous sounds in every context. Since the off-glide element has exactly the same quality as the monophthong, the dishomogeneity of [i.j] and [u.w] is hard to perceive.

The long close vowels are not allowed in word-final position, so by analogy with the half open vowels /ɛ:/ and /ɔ:/, for which the same holds, one would expect them to end schwa-like. This is what actually occurs, for it is not uncommon for /i:/ and /u:/ to be realized as [iə] and [uə] — though not by all speakers —, especially when they precede a dental-alveolar consonant, the words tiid /ti:d/ time and hûs /hu:z/ house have the 'normal' realizations [ti:t] and [hu:s], but [tiət] and [huəs] occur as well (see Graaf 1985:27, Visser 1997:24).

This means that the realization of these long vowels may coincide with that of the centring diphthongs /iə/ and /uə/ (see diphthongs in Frisian). In this case, the centring diphthongs are allophones of the long monophthongs. However, the difference between a long monophthong and the 'corresponding' centring diphthong can also be contrastive, as shown in in the tables below for the pairs /i:/ - /iə/ and /u:/ - /uə/:

Table 1: Minimal pairs with /i:/ - /iə/
/i:/ /iə/
iis /i:z/ ice ies /iəz/ bait
bliid /bli:d/ glad, happy blied(e) /bliəd/ to bleed
riid /ri:d/ channel, trench ried /riəd/ council; advice
wiid en siid /si:d/ far and wide sied /siəd/ seed
striid /stri:d/ fight; struggle stried /striəd/ struggled (past tense stem of the verb strid(e) to struggle)
Table 2: Minimal pairs with /u:/ - /uə/
/u:/ /uə/
hûd /hu:d/ skin hoed /huəd/ hat
bûd /bu:d/ bump, lump boerd /buəd/ board; blackboard
gûz(je) /ɡu:z/ to cry, to scream goes /ɡuəz/ goose

As noted before, the realization of /i:/ and /u:/ as [iə] and [uə] is not uncommon. But this does not mean that the long monophthongs are being supplanted by the centring diphthongs. The fact that the difference between the two is contrastive, as shown in the table above, is to be held responsible for this.

It cannot be denied, though, that the phonological opposition between /i:/ and /iə/ and /u:/ and /uə/ is losing ground. Graaf (1985) noted that younger speakers of Frisian had shorter realizations of the long close vowels than older speakers. Recent measurements reveal that this tendency has persevered. Nevertheless, younger speakers hold on to a clear difference between the short and long version of these vowels, so that the length distinction remains suitable for encoding a phonological contrast (see Gilbers et al. 2012). As a result of ongoing shortening, however, this appropriateness may be expected to diminish gradually, which may lead speakers to substitute the opposition long vs. short for these vowels by an opposition centring diphthong vs. short vowel.

In order to keep a symmetrical vowel system, in which each short monophthong has a non-short counterpart, some speakers are giving up the contrast between /i:/ and /iə/ and /u:/ and /uə/. The upshot of this is that the pairs of words in the table above are gradually becoming homophones, whereas minimal pairs like tyk /tik/ tick - tiik /ti:k/ mattress-cover, pillow-case and hoes /huz/ (record) sleeve, dust cover - hûs /hu:z/ house are making way for /tik/ - /tiək/ and /huz/ - /huəz/. In this way the vowel system remains symmetrical, though in a less straightforward way. Originally, there were nine pairs of vowels which showed a purely quantitative contrast: short vs. long. In the developing new system, six of these pairs are left, whereas the close vowels have entered into a contrast which is both quantitative and qualitative: short and monophthongal vs. long and diphthongal.

  • Boersma, Johannes. & Woude, Goasse van der1972Spraeklear I. Lesboekje foar de oplieding ta de Fryske AkteLjouwertAfûk
  • Gilbers, D.G, , Visser, W. & Weening, J.G2012Het verschil tussen de lange en korte hoge vocalen in het Fries en de invloed hiervan op de realisering van de Nederlandse hoge vocalen door FriestaligenUs Wurk6169-90
  • Graaf, Tseard de1985Phonetic aspects of the Frisian vowel systemNowele523-40
  • Graaf, Tseard de1985Phonetic aspects of the Frisian vowel systemNowele523-40
  • Riemersma, Tr1979Sylabysjerring, nazzeljerring, assymyljerringLjouwertKoperative Utjowerij
  • Sipma, Pieter1913Phonology and Grammar of Modern West FrisianLondon, New YorkOxford University Press
  • Visser, Willem1997The Syllable in FrisianVrije Universiteit AmsterdamThesis