We report ongoing investigations into phonetic change of dialect groups in the northern Netherlandic language area. We are particularly interested in the Frisian and Low Saxon dialect groups, because the vitality (the aggregated likelihood of propagation given internal and external factors; cf. Edwards, 1992) of these neighboring varieties differs considerably (Ytsma, 2006; Bloemhoff et al., 2013). It is known that both dialect groups are in decline, though their rate differs (Driessen, 2012). At the same time there are some further developments in the Netherlandic area, such as the formation of so called 'regiolects' (stable varieties between the standard language and traditional varieties; cf. Hoppenbrouwers, 1990; Vandekerckhove, 2009; Ghyselen, 2015), which reinforce the dialects.

We analyze existing phonetic corpora (Taeldeman & Goeman, 1996; Heeringa & Hinskens, 2014), which consist of recordings and phonetic transcriptions of older male dialect speakers translating Standard Dutch words into their own dialect. The overlapping locations and words between these datasets (24 locations, 36 words) are then used to compare the realizations of Standard Dutch target words over time. Instead of using the existing transcriptions from these datasets, however, a single transcriber made new phonetic transcriptions. This approach avoids intertranscriber variability, which is a known issue for one of the two datasets (Hinskens & Van Oostendorp, 2006). We analyze these data with dialectometric approaches that allow us to quantify change among speakers of Frisian and Low Saxon.

A three-dimensional variant of the Levenshtein distance (Heeringa & Hinskens, 2015), combined with methods that induce realistic phonetic distances between transcriptions (Wieling et al., 2012), is used to estimate how much dialect groups have changed between 1990 and 2010, and whether they changed towards Standard Dutch or away from it. Using generalized additive models (Wood, 2017), we can then investigate the non-linear spatial variation of the observed phonetic change, while we directly account for the inter-item variation between Standard Dutch target words.

Our analyses indicate that phonetic change is a slow process in this geographical area, as the phonetic change overall is low in the 20 year time span of interest (i.e., approximately 3.4% across the area). Note that there is only statistical evidence of spatial change when the Low Saxon group is contrasted with the rest of the dialect area (the Groningen dialects are excluded from the Low Saxon group as they exhibit different phonological and morphological patterns). This indicates that phonetic language change in the area is a spatially chaotic process (cf. Heeringa & Nerbonne, 2000; Heeringa & Hinskens, 2015). The Frisian and Groningen dialect groups are the least convergent towards Standard Dutch, while the Low Saxon varieties are the most prone to change of this type.

We offer possible explanations for our findings (e.g., on the basis of what is known about the language status and the respective speaker populations), and we discuss shortcomings of the data and approach. These considerations serve to inform fellow dialect researchers of the potential of our approach, but also highlight the necessary precautions when doing research of this type.