Aerodynamic and acoustic characteristics of the voicing distinction in electronic larynx speech

Abstract

Speech attributes that depend primarily on laryngeal adjustments in normal speech can also be produced by skilled alaryngeal speakers. For example, listeners reliably perceive the consonant voicing distinction in the production of good electronic larynx speakers. In most cases, electronic larynx speech is produced with a continuous sound source regardless of the voicing features of consonants. Therefore, voicing distinctions are evidently not produced by systematic control of voice timing relative to consonantal closure. It has been suggested that electronic larynx speakers use some alternative strategies, such as markedly higher intraoral pressure for voiceless consonants than for their voiced cognates, producing different acoustic patterns for different voicing features. In this study, acoustic and intraoral pressure measurements were made during multiple repetitions of nonsense syllables produced by laryngectomees and normals with and without (transcervical-type) electronic larynges. Different aerodynamic and acoustic patterns pertaining to voicing distinctions in normal and electronic larynx speech were observed. Some specific mechanisms in electronic larynx speech will be discussed. [Work supported by NIH Grant NS13870.]