Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Cross-Linguistic Comparison of Stress Processing in Familiar Words Among 2- to 5-Year-Olds: Languages With Contrastive (German) and Without Contrastive Stress (Zurich German)
Poster Session B, Wednesday, September 30, 4:30 - 6:30 pm, Wangari Maathai
Jana Engel1, Claudia Friedrich1, Ulrike Schild1; 1University of Tuebingen
The informational value of word-level stress – expressed through stressed and unstressed syllables – varies across languages. In variable-stress languages such as German, a stressed syllable can occur in different positions within a word. In fixed-stress languages, by contrast, stress is realized in the same syllable position in every word. In Zurich German, for example, stress is typically word-initial, with only few exceptions. Adults with a variable-stress language use stress cues for word recognition as soon as they can extract them from the word. Children acquiring these languages also use stress cues in word recognition by the end of the first year of life. For word recognition in a language with fixed stress, stress is less informative. Indeed, adults adjusted to those languages tend to be less sensitive to stress cues than adults adjusted to variable-stress languages are. Here, we tested whether sensitivity to stress cues for word recognition develops differently in children acquiring a variable-stress language (German) vs. a fixed-stress language (Zurich German). We examined stress processing in early-acquired words in children aged between 2;0 and 4;11 years learning either German (N = 45) or Zurich German (N = 40). Adult speakers of German (N = 20) and Zurich German (N = 21) served as controls. We expected stress effects in German-speaking adults and children because stress cues are informative for lexical processing in German. In contrast, Zurich German-speaking adults and children should show weaker or no stress effects. Further, children’s stress processing should increasingly adapt to the prosodic structure of their native language with age, such that stress effects would increase in the German group but decrease in the Zurich German group. We used a modified looking-while-listening paradigm, in which participants viewed one novel and one familiar object while hearing either a correctly or incorrectly stressed version of the familiar object’s name. German-speaking adults showed higher fixation proportion (PI) to the novel object in incorrectly stressed trials and corresponding time-course differences in growth curve analysis (GCA), indicating reliance on stress during word recognition. Against expectations, German-learning children showed no stress effect in PI and only a small stress effect in the GCA. Notably, the GCA effect decreased with age. Zurich German-speaking adults showed no stress effect in PI, although the GCA indicated small stress-related differences in the time course of looking. Zurich German-learning children likewise showed no effect in PI. However, the GCA revealed an interaction between stress and age, such that looking to the novel object in incorrectly stressed trials decreased with age. Together our findings suggest higher sensitivity to stress variation in children and adults adjusted to German than in children and adults adjusted to Swiss German. However, children with both target languages increasingly tolerated prosodic variation between the age of two to four years. This developmental pattern was in line with our hypothesis for Zurich German, a fixed-stress language, but contrary to our expectation for German, a variable-stress language. We discuss potential different aspects of language acquisition related to this effect for both groups.
Topic Areas: Language Development/Acquisition, Prosody