Phonetic Detail Encoding in Explaining the Size of Speech Planning Window

Shan Luo


With the ultimate goal of understanding the production planning scope, this study manipulates phonetic information (place of articulation and voicing) and measures three acoustic cues to analyze consonant clusters across words produced by English (L1) and Mandarin (L2) speakers. We continue to explore a) how phonetic detail interacts with prosodic boundary in modulating surface realization, and b) the roles of phonetic information in speech planning motor control. The results show that L2 speakers exhibited different acoustic deviations varying with their proficiency level. The group with lower L2 proficiency significantly deviated from the L1 group in release likelihood and closure shortening, while the higher-proficiency group exhibited less nativelike performance in terms of closure durations. The results also discover that all speakers are subject to language-independent articulatory constraint at word boundaries, while language-specific phonetic detail accounts for more nonnative deviations. The core findings highlight a long-distance speech planning scope in native speech, with cross-word phonetic information interacting with prosodic encoding. It is argued that phonology applies blindly across words and is independent of lexical cognitive load.


 DOI: 10.21437/Interspeech.2019-1412

Cite as: Luo, S. (2019) Phonetic Detail Encoding in Explaining the Size of Speech Planning Window. Proc. Interspeech 2019, 1766-1770, DOI: 10.21437/Interspeech.2019-1412.


@inproceedings{Luo2019,
  author={Shan Luo},
  title={{Phonetic Detail Encoding in Explaining the Size of Speech Planning Window}},
  year=2019,
  booktitle={Proc. Interspeech 2019},
  pages={1766--1770},
  doi={10.21437/Interspeech.2019-1412},
  url={http://dx.doi.org/10.21437/Interspeech.2019-1412}
}