Sabine Gosselke Berthelsen, University of Copenhagen
Prosody is important for speech processing, maybe particularly so in the Scandinavian languages where word tones and voice quality play an important role in differentiating words and cueing suffixes. In Danish, for instance, a creaky voice on a word stem (˷) can differentiate a thought from a petrol station (tanke-n, /tsaŋk.ən/, ‘the thought’ vs tank-en, /tsaŋ̰k.ən/, ‘the petrol station’) but it can also cue grammar and prepare the listener, for example, for a plural or singular suffix (pil-e, /pʰiːl.ə/, ‘arrows’ vs pil-en, / pʰḭːl.n/, ‘the arrow’ ). In Swedish and Norwegian, the same contrasts are expressed with tones. This systematic interaction between prosody and inflections allows for facilitated speech processing in the Scandinavian languages (Clausen & Kristensen, 2015; Roll et al., 2017). We propose, however, that the facilitative function of prosody is only available to native speakers and second language learners at high proficiency levels and we, therefore, investigate whether training could benefit L2 learners at earlier stages.
In neurophysiological and behavioural studies, we compared second language learners’ awareness of and responses to L2 prosody to those of native speakers of Danish and Swedish. Specifically, we presented inflected words where the prosody either correctly or incorrectly predicted upcoming suffixes and asked participants to make grammatical decisions based on the suffix. We also presented minimal pairs, only distinguished by prosody, and asked participants to indicate the correct word. Our results indicate that native speakers easily distinguish minimal pairs and were disrupted in their suffix processing when the prosody was mismatching. Second language learners were not affected by mismatched prosody and could not distinguish minimal pairs based on an unfamiliar prosodic feature. A specifically devised training paradigm, however, has provided promising results for improved processing of L2 prosody. Thus, beginner learners’ distinction of minimal pairs quickly improved and we observed facilitated processing after training, i.e., anterior negativities indicative of grammar processing. We also observed more native-like processing if L2 prosody was similar to L1 prosody.
References
Clausen, S. J., & Kristensen, L. B. (2015). The cognitive status of stød. Nordic Journal of Linguistics, 38(2), 163-187. doi:10.1017/S0332586515000141
Roll, M., Söderström, P., Frid, J., Mannfolk, P., & Horne, M. (2017). Forehearing words: Pre-activation of word endings at word onset. Neuroscience Letters, 658, 57-61. doi:10.1016/j.neulet.2017.08.030