Phonetics, Phonology and the Central Franconian Tone

Carlos Gussenhoven, University of Nijmegen (

The conceptual distinction between phonetic implementation and the phonological grammar may become particularly clear from phonological change. In the typical case, speakers modify their phonetic implementation of a given phonological element, sometimes progressively over several generations, until a new generation of speakers constructs a different representation from that which previous generations constructed, one which more directly reflects the new phonetic form. Crucially, the motivation for the changed phonetic behavior is independent of the motivation for choosing a particular representation.

The series of phonological changes that began with the introduction of a lexical tone in the Low German of Cologne around 1300 presents an interesting illustration of this independence. Small phonetic changes motivated by articulatory and perceptual efficiency led to new phonological representations motivated by cognitive efficiency and consistency, ultimately giving rise to grammars like that of Roermond Dutch, which is atypical within West Germanic.

The motivation for the phonetic change that led to the introduction of the lexical tone was a need to lengthen vowels in singular monosyllabic nouns which had phonologically long vowels in their plurals, a need that was absent in the precursor of standard Dutch, in which the short vowels were preserved (dal - dalen, slot -sloten, etc.). Since in the Central Franconian dialect the plural forms were also monosyllabic, an interpretation of the phonetically lengthened vowel in the singular as a phonologically long vowel was excluded, resulting in a tonal representation of the singulars, Accent 2, by the side of toneless plural forms, Accent 1. Two subsequent segmental changes are (a) the split of diphthongs into more and less diphthongal pronunciations in syllables with Accent 1 and Accent 2, respectively, and (b) the split of mid vowels into opener and closer vowels in syllables with Accent 1 and Accent 2, respectively. I will argue that both changes are to be explained as attempts by speakers to make syllables with Accent 2 sound longer. I will present perceptual data to show that closer vowels sound longer than opener vowels, and will explain this fact by appealing to a general theory of compensatory listening.

CLCG Klankleer Group Workshop 'On the Boundaries of Phonology and Phonetics'