Pauses and Speech Planning

alt_text 

We found support for the hypothesis that pauses at major syntactic boundaries (i.e., grammatical pauses), but not ungrammatical (e.g., word search) pauses, are planned by a high-level cognitive mechanism that also controls the rate of articulation around these junctures.

We used real-time magnetic resonance imaging to analyze articulation at and around grammatical and ungrammatical pauses in spontaneous speech. We found that grammatical pauses had an appreciable drop in speed at the pause itself as compared to ungrammatical pauses, which is consistent with our hypothesis that grammatical pauses are indeed choreographed by a central cognitive planner.

For more details, please refer to the following publication:

  • Vikram Ramanarayanan, Erik Bresch, Dani Byrd, Louis Goldstein and Shrikanth S. Narayanan (2009), Analysis of pausing behavior in spontaneous speech using real-time magnetic resonance imaging of articulation, in: Journal of the Acoustical Society of America Express Letters, 126:5 (EL160-EL165) [link].

Planning and Execution in Different Uses of the Vocal Instrument

alt_text 

This work examined three different behavioral modalities of production in the human vocal tract: read speech, spontaneous speech and singing. We observed and quantified differences in planning and execution of these three behaviors elicited from 4 formally trained, professional sopranos. To analyze the data, we proposed fully automatic measures of average articulator speed, posture and acoustic spectra. We found evidence that these different behavioral modalities involve speech planning mechanisms in different ways (see table on left).

For more details, please refer to:

  • Vikram Ramanarayanan, Adam Lammert, Dani Byrd, Louis Goldstein and Shrikanth Narayanan (2011). Planning and Execution in Soprano Singing and Speaking Behavior: an Acoustic–Articulatory Study Using Real-Time MRI, in: International Speech Production Seminar 2011, Montreal, Canada, June 2011 [pdf].

Lexical Class, Word Frequency and Articulatory-Acoustic Covariation

alt_text 

Word frequency and lexical class distinction between function and content words have been shown to significantly influence word production (see figure on left). In this work, we used real-time magnetic resonance imaging to investigate the effect of word frequency and lexical class on articulatory characteristics (the articulator speed) as well as acoustic characteristics (pitch and short-term energy) in word production.

A Granger causality analysis uncovered a causal relationship from articulatory speed to pitch/energy for low-frequency content words. We further observed, using functional canonical correlation analysis, a tight covariation of articulatory and acoustic characteristics for low-frequency content words. These results support the view that word frequency distinctly influences the production of function and content words as manifested in their articulation and acoustics, as well as the dynamic coupling of these temporal streams.

For more details, please refer to:

  • Zhaojun Yang, Vikram Ramanarayanan, Dani Byrd and Shrikanth Narayanan (2013). The effect of word frequency and lexical class on articulatory-acoustic coupling, in proceedings of: Interspeech 2013, Lyon, France, Aug 2013 [pdf].