The intelligibility of tracheoesophageal speech, with an emphasis on the voiced-voiceless distinction

Abstract

Total laryngectomy has far-reaching effects on vocal tract anatomy and physiology. The preferred method for restoring postlaryngectomy oral communication is prosthetic tracheoesophageal (TE) speech, which like laryngeal speech is pulmonary driven. TE speech quality is better than esophageal or electrolarynx speech quality, but still very deviant from laryngeal speech. For a better understanding of neoglottis physiology and for improving rehabilitation results, study of TE speech intelligibility remains important. Methods used were perceptual evaluation, acoustic analyses, and digital high-speed imaging. First results show large variations between speakers and especially difficulty in producing voiced-voiceless distinction. This paper discusses first results of our experiment.

Quality of ‘glottal’ stops in tracheoesophageal speakers

“…On the other hand, Lü ndstrom and Hammarberg (2004) used high-speed imaging to investigate speech in four TE speakers and found that all four speakers were able to make opening gestures in the neoglottis while producing voiceless stops. Jongmans et al (2006) used videokymography to illustrate the difference between two speakers when producing the sequence 'patakabadaga'. One speaker was unfortunately unable to produce any vibrations, but the other was able to stop vibrations when producing the voiceless consonants.…”

“…It is thought that the closure and tonicity, combined with the amount of speakers' control over the neoglottis, have consequences for the production of the voiced-voiceless contrast. Regarding closure and tonicity, Jongmans, Hilgers, Pols, and Van As-Brooks (2006) used digital high-speed imaging to investigate the production of the voiced-voiceless distinction. From these images of the neoglottis, they deduced that speakers with a hypotonic, incomplete neoglottic closure were more likely to produce voiced consonants as voiceless, whereas speakers with a closed neoglottis were more likely to produce voiceless consonants as voiced.…”

Using Beatboxing for Creative Rehabilitation After Laryngectomy: Experiences From a Public Engagement Project

“…There are four different ways to restore speech, with TE voice among them nowadays considered to be the gold standard (Jongmans et al, 2006;Elmiyeh et al, 2010;Bohnenkamp et al, 2011).…”

“…Overall, this approach produces a more natural sound, requires greater power from the lungs, facilitates more volume control (van As et al, 1998;Jongmans et al, 2006;Kaye et al, 2017), makes it easier to differentiate between voiced and unvoiced consonants (Jongmans et al, 2006), is expensive and necessitates that the valve be replaced every 3-6 months (Staffieri et al, 2006), and offers patients a higher degree of voice satisfaction (Kaye et al, 2017).…”

“…One of the difficulties experienced in speech after laryngectomy, in addition to the issues of volume control and pitch range, is the differentiation between voiced and unvoiced consonants (Jongmans et al, 2006). The variation between these consonants simply depends on whether or not the voice is used to support the articulation.…”