Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Prosodic phrasing in six-month-old Dutch-learning infants
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster E37 in Poster Session E, Thursday, October 26, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Rachida Ganga1, Frank Wijnen1, Aoju Chen1; 1Institute for Language Sciences, Utrecht University
Introduction: The sentence ‘Miffy says Snuffy is my best friend‘ may leave you puzzled as to who declares the friendship status to whom. In West-Germanic languages, speakers can resolve the ambiguity by inserting an intonational phrase (IP) boundary after ‘Miffy says’ or surrounding ‘says Snuffy’. Such boundaries can be marked by three types of prosodic cues: a rise in pitch and lengthening of the IP-final syllable, and a pause following the IP. Behavioral research shows that infants need fewer cues to recognize IP boundaries as they age, e.g., English-learning infants distinguish fully-marked boundaries at four months and only need pitch rise and another cue at six months. Similarly, six-month-old German-acquiring infants become less dependent on pauses at eight months. An EEG study on German-learning infants’ IP boundary processing, reflected in the ERP component Closure Positive Shift (CPS), showed that they rely on final lengthening in the absence of pauses and do so already at six months. The change in infants’ sensitivity to IP boundary cues depends on the relative importance of the cues in the native language. English speakers rely mostly on pitch rise but German speakers on final lengthening. Past research suggests that Dutch speakers mostly use pauses. However, recent findings show that Dutch adult speakers mostly rely on final lengthening to produce IP boundaries, similar to German speakers. This raises the general question whether Dutch-learning infants acquire IP boundaries similarly to German-learning infants. Specifically, we have addressed the following questions: 1) Do six-month-old Dutch-learning infants show CPS after hearing two-cue boundaries, and 2) is the CPS larger after fully-cued boundaries? || Methods: Thirteen Dutch-learning infants (mean age: 6m21days) listened to sequences of three names connected by ‘and’ (Dutch: en). Adopting the paradigm of the German ERP study, the stimuli were recorded by a female speaker of Dutch and manipulated to create three conditions: A) without a boundary, B) a boundary cued by pitch rise and final lengthening after the second name, and C) a fully marked boundary after the second name. ERPs were computed from the acoustic onset of the boundary in four scalp regions (left and right frontal and posterior, LF; RF; LP; RP respectively). || Results: Linear mixed models revealed that the two-cue boundary in comparison to the no-boundary elicited global positivities in the RF, LP, and RP regions (p<0.001, p=0.021, and p=0.001 respectively) and a negative shift in the LF region (p<0.001). The fully-marked boundary resulted in larger positivities than the two-cue boundary in the RP region (p<0.001) and in less positivity in the right frontal region (p<0.001). || Conclusion: Six-month-old Dutch-learning infants processed two-cue IP boundaries as shown by a broad positivity, which we interpret as the CPS. A fully-marked boundary resulted in a stronger CPS in the RP region, but weaker in the RF region. These findings suggest that Dutch-learning infants are similar to their German-learning peers and can already process IP boundaries in the absence of the pause cue at the age of six months.
Topic Areas: Prosody, Language Development/Acquisition