Lange en koarte ienlûden yn it Frysk: in reaksje op De Haan

Abstract

In most of the literature, it is assumed that Frisian has a symmetrical vowel system, consisting of nine short and nine corresponding long vowels, see (i) (schwa omitted):

           (i)      the traditional symmetrical vowel system

                   short vowels:        a, ε, ɔ, I, ö, o, i, y, u

                   long vowels:         a:, ε:, ɔ:, e:, ö:, o:, i:, y:, u:

 The short vowels are assumed to occupy two structural positions, the long vowels one.

     This traditional view has been challenged by De Haan in his 1999 article “Frisian monophthongs and syllable structure” (Us Wurk, 48, pp. 19‑30), in which he arrives at the asymmetrical classification in (ii):

(ii):the asymmetrical system according to De Haan (1999)

     A‑vowels:             i, i:, y, y:, u, u:, e:, ö:, ε:, o:, ɔ:, a:

     B‑vowels:             I, ö, ε, o, ɔ, a

Here, the A‑vowels - including /i/, /y/, and /u/ - are assumed to occupy two structural positions, the B‑vowels one.

   De Haan adduces several distributional arguments for this new classification, some of which are familiar from the phonological literature: 1) the A‑vowels must occur in an open syllable, the B‑vowels in a closed one; 2) the A‑vowels may occur word‑finally, the B‑vowels must not; 3) 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; 4) the velar nasal /ŋ/ can only be preceded by a B‑vowel; 5) there is a tendency for A‑vowels to precede a voiced fricative, whereas B‑vowels tend to be followed by a voiceless fricative. In this article, these arguments are, first of all, confronted with a fair number of counter­exam­ples, which diminishes their force.

But there is also other evidence to the contrary. The vowels /i/-/i:/, /y/-/y:/, and /u/-/u:/ show up in minimal pairs. Since they occupy the same number of structural positions, viz. two, they must be distinguished on the basis of an independently motivated phonological feature. De Haan suggests that the features ‘lax’ vs. ‘tense’ might apply here, although he does not elaborate on this. However, invoking lax-tense as a fundamental qualitative distinction in Frisian phonology meets with several objections, which the (quantitative) distinction short‑long does not.

     De Haan’s classification also runs into problems as to a proper char­ac­terisation of falling and centring diphthongs and as to a proper under­standing of their distributional behaviour. Since these diphthongs show the same distributional properties as long vowels, they are assumed, also by De Haan, to occupy two structural positions. But since they consist of the combination of a B‑vowel with one of the A‑vowels /i/, /y/, or /u/, they would have to occupy three structural positions. This is at odds with the general assumption that the syllable nucleus should not oc­cu­py more than two structural positions, whereas the fact that these diphthongs may be followed by non‑coronal consonants in word‑final position also points to a bipositional constellation.

     All in all, there seems little reason to adopt De Haan’s proposal.