Evaluation of prosodic juncture strength using functional data analysis.

Area of Science:

• Phonetics
• Speech Science
• Computational Linguistics

Background:

• Prosodic structure significantly influences speech timing by shaping articulatory events.
• Quantifying prosodic effects on articulation is crucial for understanding speech variability and generalizing findings across speakers.

Purpose of the Study:

• To present a novel, automated method for quantifying prosodic boundary strength using articulatory speech data.
• To introduce the Deformation Index, a new variable derived from functional data analysis (FDA).

Main Methods:

• Utilized functional data analysis (FDA) on articulatory speech data.
• Proposed and calculated the Deformation Index (area under FDA time-deformation functions).
• Validated the method using synthetic speech (TaDA model) and real speech from four speakers.

Main Results:

• The Deformation Index successfully captured known differences in boundary strengths from synthetic speech.
• The method accurately differentiated boundary types in non-synthetic speech data.
• Demonstrated generalizability across different speakers and linguistic contexts.

Conclusions:

• The Deformation Index provides a robust and generalizable measure of prosodic boundary strength from articulatory data.
• This method offers a systematic way for speech scientists to quantify prosodic influences on articulation.
• The findings contribute to a deeper understanding of speech timing and prosody.