• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
Long and short monophthongs: a different view
quickinfo

Instead of the symmetrical system of nine short and nine long vowels (monophthongs), which is generally adopted in Frisian phonology, De Haan (1999) proposes an asymmetrical system. He divides the vowels into two classes, the A-vowels and the B-vowels, which differ with respect to the number of structural phonological positions they occupy: two (the A-vowels) vs. one (the B-vowels). This topic describes this asymmetrical system and argues that it has its advantages and disadvantages.

readmore

According to most literature, Frisian has a symmetrical vowel system of nine short and nine corresponding long vowels, see the table below (schwa omitted):

Table 1: The symmetrical vowel system according to tradition
short vowels: a, ɛ, ɔ, ɪ, ø, o, i, y, u
long vowels: a:, ɛ:, ɔ:, e:, ø:, o:, i:, y:, u:

This traditional view has been challenged by Haan (1999), who arrives at the asymmetrical classification in the table below:

Table 2: The asymmetrical vowel system according to Haan (1999):
A-vowels: i, i:, y, y:, u, u:, e:, ø:, ɛ:, o:, ɔ:, a:
B-vowels: ɪ, ø, ɛ, o, ɔ, a

De Haan (1999) divides the vowels into two classes, the A-vowels and the B-vowels, a terminology which is neutral as to vowel length. The class of A-vowels not only comprises long vowels, but also the three short vowels /i/, /y/, and /u/. This means that all close vowels, whether they are long or short, belong to this class. An A-vowel is assumed to occupy two structural phonological positions, a B-vowel only one.

De Haan adduces several distributional arguments for the above classification, some of which are familiar from the phonological literature:

  • the A-vowels must occur in an open syllable, the B-vowels in a closed one;
  • the A-vowels may occur word-finally, the B-vowels must not;
  • in word-final position, the A-vowels can be followed by no more than one non-coronal consonant, the B-vowels can be followed by two such consonants;
  • the velar nasal /ŋ/ can only be preceded by a B-vowel;
  • there is a tendency for A-vowels to precede a voiced fricative, whereas B-vowels tend to be followed by a voiceless fricative.
It should be noted that these distributional statements have quite a number of exceptions Visser (2003).

There are two other pieces of evidence that De Haan's approach is on the right track. In the first place, there is orthographical evidence. Frisian orthography makes use of doubling of vowel symbols in order to denote long monophthongs, but only in closed syllables, see a minimal pair like lam /lam/ lame ~ laam /la:m/ lamb. Since there are too few vowel symbols to distinguish all pairs of short and long monophthongs in an unambiguous way, there is doubling of consonant symbols in case the latter stand between a symbol representing a short monophthong and schwa. Consonant doubling thus is used as a means to indicate the shortness of a (preceding) monophthong. It pertains to the B-vowels, which are written as <i> ( /ɪ/), <u> ( /ø/), <e> ( /ɛ/), <o> ( /o/), <o>/<a> ( /ɔ/), and <a> ( /a/). This practice is exemplified in the table below:

Table 3: Examples of consonant doubling between a B-vowel and schwa
tille /tɪlə/ small fixed (wooden) bridge
mudde /mødə/ hectolitre
fekke /fɛkə/ shrew
rosse /rosə/ to rub (infinitive; all plural persons present tense)
kanne /kɔnə/ jug
goffe /ɡɔfə/ cack-handed person
karre /karə/ cart

Consonant doubling is not found after the vowel symbols <i> ( /i/), <u> ( /y/), and <oe>/<û> ( /u/), though these denote a short vowel; see the examples in the table below:

Table 4: Examples of non-doubled consonant symbols between a short A-vowel and schwa
side /sidə/ side; page; silk
rude /rydə/ scabies
tûke /tukə/ branch
boeken /buk+ən/ books

When the official spelling of Frisian was decided on — as late as 1879 — there must have been some sort of understanding of the difference between the short A-vowels on the one and the real short vowels, the B-vowels, on the other hand.

extra

In fact, the last two examples of the table above, viz. tûke and boeken, give a distorted picture of doubling. A consonant symbol is never doubled after a vowel symbol with a diacritic (tûke) or after a complex vowel symbol (boeken), unless a complex vowel symbol denotes a (short) rising diphthong, as in stiennen [stjɪnn̩] stones (cf. stien [stiən] stone), sleatten [sljɛtn̩] (cf. sleat [slɪət] ditch), fuotten [fwotn̩] feet (cf. foet [fuət] foot), and soannen [swann̩] sons (cf. soan [soən] son). The latter links up with the fact that the two complex consonant symbols of Frisian orthography, viz. <ch> and <ng> — denoting /x/ and /ŋ/, respectively — are not doubled either, as illustrated by richel /rɪxəl/ ledge, ridge (*richchel) and tonge /toŋə/ tongue (*tongnge).

Secondly, there is evidence pertaining to the process of Homorganic Glide Insertion (see the resolution of vocalic hiatus in general). This takes place between a full vowel and schwa, but only if the former is long, as it is in seeën /se:+ən/ [se:jən] seas and dowen /do:+ən/ [do:wən] pigeons. When an unsuspected B-vowel like /ɪ/ shows up in such a configuration of vocalic hiatus, glide insertion is accompanied by vowel lengthening, as in breaën /brɪə+ən/ [brɪ.jən] rye breads. Underlyingly, there are two two adjacent schwas here, one stem-final and one suffix-initial. Due to degemination (see the process of degemination), one of these remains. It should also be noted that the vocalic portion of both the stem and the suffix is preserved, for breaën is a disyllabic form, which is not in rhyme with the underived word lean /lɪən/ wages. The short close vowels /i/ and /u/, however, trigger glide insertion, without having to undergo lengthening, as in dyen /di+ən/ those ones and wy hoeë dat net te dwaan /hu+ə/ [huwə] we need not do that (hoe /hu/ is the shortened stem of hoech/hoef /hu{ɣ/v}/ need not). This is indicative of /i/ and /u/ behaving as long vowels (A-vowels).

But there is also evidence against the bipartitioning of the vowel system above. The vowel pairs /i/ - /i:/, /y/ - /y:/, and /u/ - /u:/ figure in minimal pairs (see long and short monophthongs). Since these are all A-vowels, they occupy the same number of structural positions, viz. two. This means that the contrast between them must be due to an independent phonological feature. De Haan suggests that the feature pair lax-tense might do the job here, although he does not elaborate on this.

Invoking lax-tense as a fundamental qualitative distinction in Frisian phonology meets with the general problem that 'tense' is a feature with no clear phonetic correlate. According to Lass (1984:92)there are no qualities attributable to 'tense' that cannot be reduced to the traditional dimensions of height, backness and duration. Ladefoged (1993) describes 'tense' as a term with no specific phonetic correlates, used when dividing vowels into classes on phonological grounds. It should be noted that the difference between short and long monophthongs in Frisian is clearly perceptible and that it is confirmed by phonetic measurements.

If /i/, /y/, and /u/ both occupy two structural positions and are characterized as 'lax', this goes against the general assumption that bipositionality links up with +tense and monopositionalness with -tense.

De Haan's classification runs into problems if it comes to a proper characterization of the complex nuclei of Frisian (see diphthongs in Frisian). Take the falling diphthongs, e.g. /ɛi/ and /ɔu/, as in reid /rɛid/ reeds and gau /gɔu/ quick, fast. Such diphthongs have the distributional properties of long vowels, hence they are expected to occupy two structural positions. Within the asymmetrical vowel system, however, /ɛi/ and /ɔu/ occupy three such positions, for they are the combination of a B-vowel, viz.  /ɛ/ or /ɔ/, and an A-vowel, viz. /i/ or /u/. Occupying three structural positions is at odds with the general assumption that the syllable nucleus may occupy no more than two. Moreover, such diphthongs may be followed by non-coronal consonants within the same stem, which also points to the bipositionality of diphthongs (the size of the word-internal and word-final syllable rhyme).

References
  • Haan, Germen J. de1999Frisian monophthongs and syllable structureUs Wurk4819-30
  • Haan, Germen J. de1999Frisian monophthongs and syllable structureUs Wurk4819-30
  • Haan, Germen J. de1999Frisian monophthongs and syllable structureUs Wurk4819-30
  • Ladefoged, Peter1993A course in phoneticsFort WorthHarcourt Brace Jovanovich College Publishers
  • Lass, Roger1984Phonology: An introduction to basic conceptsCambridgeCambridge University Press
  • Visser, Willem2003Lange en koarte ienlûden yn it Frysk: in reaksje op De Haan (1999)Us Wurk52130-154
printreport errorcite