Ghost segments in the Flemish Tussentaal
|Title||Ghost segments in the Flemish Tussentaal|
|Year of Publication||2022|
|Conference Name||Dag van de Fonetiek 2022|
|Publisher||Nederlandse Vereniging voor Fonetische Wetenschappen|
|Conference Location||Utrecht, The Netherlands|
The Flemish informal standard (or Tussentaal) exhibits a feature present in most Low Franconian variants of Flanders: the elision of final coronal plosives in words such as wat, niet and met, whereby they are realised without their final stop: [βɑ], [ni] and [mɛ]. Notably, these words participate in a sandhi process that triggers the devoicing of any fricatives that follow, similar to how those fricatives would devoice if the elided coronal were still present. Thus, phrases such as <wat vlees> and <met zand>, in which the second word normally starts with a voiced fricative, are realised as [βɑ fle:s] and [mɛ sɑnt]. The elision of this final coronal happens in some words (met, gaat, wat) and not in others (wet, tot, uit). While previous literature has acknowledged the existence of this ghost segment (such as Camerman, 2007; De Schutter, 1999; Rys, 2020), there had been little to no synchronic or diachronic account for its distribution prior to this study. The present study therefore set off to do the following: (1) document the presence of the feature in the Tussentaal, and (2) attempt to account for its distribution.
Using a Python web-scraping script and the corpus search engine OpenSonar (Oostdijk et al., 2018), up to (where possible) ten audio tokens each of the fifty most common Dutch words according to A Frequency Dictionary of Dutch (Tiberius & Schoonheim, 2014) that contain final -t or -d were collected from the Corpus Gesproken Nederlands (Eerten, 2007), a spoken corpus of Dutch. The scraped data pool was filtered to include only speakers of Belgian Dutch. The joint results of an algorithmic approach using Praat and a manual annotation approach reveal that final coronals are phonetically elided post-vocalically in the present tense inflectional morpheme /-t/ and most function words. All exceptions found were words that have undergone a historical process of word-final schwa deletion, suggesting that final coronal plosive deletion may be older. And indeed, evidence of final coronal plosive deletion was found in texts written in the 13th century, before word-final schwa deletion had taken place.
Camerman, Filip. (2007). Antwerps schrijven: Spelling en grammatica van het 21e-eeuwse Antwerps (pp. 44-46). de Vries-Brouwers.