Vol. 49:1 (2023) ► pp.37–69
Effects of language familiarity, utterance length, and speech quality in prosodic boundary identification
This study investigates the effects of several stimulus sources: language familiarity, utterance length, and speech quality, on listeners’ predictions of the sizes of the upcoming prosodic boundaries. Experiments with native Taiwanese speakers were conducted, and the stimuli varied in prosodic boundary units (i.e., word, phrase, and sentence), languages (i.e., Taiwanese, English, and Swedish), utterance lengths (i.e., 2-second, and 2 syllables; the latter is approximately 0.416-second long), and speech qualities (i.e., normal speech, low-pass filtered speech). Results showed that: (a) listeners performed better when the utterances were longer; (b) listeners performed better in low-pass filtered speech when they had no prior knowledge of the target language; (c) there was a tendency for the language familiarity effect, but this effect was heavily influenced by the extent of similarity of the phonetic realizations in different prosodic boundaries, and the listeners’ language proficiency which was associated with working memory storage capacity.
Article outline
- 1.Introduction
- 2.Experiment 1: Prosodic boundary identification experiment with normal
stimuli
- 2.1Method
- 2.1.1Participants
- 2.1.2Materials
- 2.1.3Procedure
- 2.1.4Data analysis and prediction
- 2.2Results
- 2.2.1The utterance length effect
- 2.2.2The language familiarity effect
- 2.2.3The prosodic boundary unit pattern
- 2.3Summary
- 2.1Method
- 3.Experiment 2: Prosodic boundary identification experiment with filtered
stimuli
- 3.1Method
- 3.1.1Materials
- 3.1.2Participants and procedure
- 3.2Results
- 3.2.1The utterance length effect
- 3.2.2The language familiarity effect
- 3.2.3The prosodic boundary unit pattern
- 3.2.4Speech qualities
- 3.3Summary
- 3.1Method
- 4.Relationship between the correctness and the prosodic cues and the relationship
between the correctness and the individual difference measures
- 4.1Predicting correctness from prosodic cues
- 4.2Predicting correctness from individual difference measures
- 5.General discussion
- 6.Conclusion
- Acknowledgments
-
References
https://doi.org/10.1075/consl.22029.kuo