Dutch has a cross-linguistically very unusual contrast in labio-dental fricatives, namely one between a fully voiced [V], a voiced and simultaneously fricated [v], and a voiceless [f]. Phonologically, these sounds are categorised as voiced glide and voiced and voiceless fricatives, respectively (Booij 1995, Gussenhoven 1999). Minimal triplets illustrating the contrast are given in (1).
(1) wee vee fee 'ache, cattle, fairy'; wijl vijl feil 'while, rasp, error'
German learners of Dutch have major problems acquiring this contrast since their native language differentiates only a voiced and a voiceless labiodental fricative ([v] and [f], respectively). German speakers usually do not distinguish the Dutch glide and voiced fricative but collapse both into one category corresponding to their native voiced fricative. This is due to the fact that both Dutch categories show vocal fold vibration, which is the only cue to distinguish German [v] from the voiceless [f] (Jessen 1998). The present study investigates the acquisition of the Dutch labio-dental fricative system by two groups of German learners of Dutch, ten in an early stage of their acquisition, and ten Germans proficient in Dutch (living in the Netherlands for more than 4 years). For this purpose, a Dutch native speaker with a three-way contrast was recorded reading five times a list that contained ten minimal triplets as in (1) and 40 filler triplets, embedded in a carrier sentence. The resulting items were randomised and auditorily presented to the subjects, who had to click on one of the orthographic representations of the three possible answer words on a computer screen. After the perception experiment, the subjects had to read the same list as the Dutch native speaker. Preliminary results indicate that proficient speakers of Dutch can distinguish better between voiced fricative and glide than beginners, but only a very small number of proficient speakers can actually produce the Dutch three-way contrast consistently.