Summary: A new study examines whether people with musical training are better at understanding vocoded speech—a degraded form of speech that models what cochlear implant users hear—than those without musical experience.
Source: Acoustical Society of America.
Cochlear implants are widely used to treat sensorineural hearing loss by electrically stimulating auditory nerve fibers via an electrode array placed inside the cochlea. While these devices can restore access to sound for many users, the resulting speech signal is often spectrally degraded and more difficult to understand than natural speech. Researchers commonly use vocoded speech—an acoustically distorted signal that simulates cochlear implant processing—in laboratory studies to investigate how listeners extract linguistic information from impoverished auditory input.
Investigators Kieran E. Laursen, Sara L. Protko, and Terry L. Gottfried from Lawrence University, together with collaborators Iain C. Williams and Tahnee Marquardt of the University of North Carolina at Wilmington and the University of Oxford, presented research on the relationship between musical experience and perception of vocoded speech at the 174th Meeting of the Acoustical Society of America (December 4–8, 2017) in New Orleans, Louisiana.
Musical ability—encompassing skills such as playing an instrument, discerning pitch and timbre, recognizing rhythmic patterns, and processing melody—has been linked in prior work to a range of cognitive and communicative advantages. Research has suggested associations between musical training and improved speech perception in noisy environments, enhanced phonetic discrimination, and benefits for learning second-language speech contrasts.
“We set out to determine whether a person’s musical background affects how they perceive vocoded speech,” Laursen explained. “The core question is whether musical expertise facilitates the detection and interpretation of pitch, intonation, and rhythm when the speech signal is acoustically degraded.”
Gottfried noted that vocoded speech preserves some temporal and rhythmic cues even as it strips away fine spectral detail. “The acoustic structure of vocoded speech differs substantially from natural speech obscured by background noise,” he said. “Because rhythmic patterns are often maintained in vocoded stimuli, musicians—who typically have extensive rhythm training—might be at an advantage. At the same time, the severe spectral loss introduced by vocoding could reduce any potential benefit from musical experience.”
The study used a commercially available experimental platform to present stimuli and record responses. Participants, both musicians and nonmusicians, listened to vocoded sentences and isolated words and were asked to transcribe what they heard. After an initial assessment, listeners were assigned to brief training sessions using either vocoded speech or natural speech, followed by another round of vocoded transcription tasks. The initial analyses found no statistically significant advantage for musicians over nonmusicians when interpreting vocoded speech, although the authors point out that limited variability in musical ability across the tested sample may have affected the outcome.
“Both groups scored well above chance on the Musical Ear Test,” Gottfried added, “so it’s possible that including listeners with very low musical aptitude would reveal differences that did not appear in this sample.” He emphasized that even null or mixed outcomes contribute valuable information about the limits and nature of any relationship between musical training and degraded-speech perception.
The implications of this work extend beyond the laboratory task of understanding vocoded sentences. Everyday speech comprehension in noisy or reverberant settings depends heavily on temporal and rhythmic cues, and acoustic challenges in those environments share features with the reductions introduced by vocoding. If musical experience were shown to improve understanding of vocoded speech, that same skill set might support better communication in real-world noisy situations for listeners with normal hearing or with hearing devices.
At the same time, the research highlights important considerations for future studies: the need for larger, more diverse samples that include a wider range of musical ability; careful characterization of what constitutes “musician” and “nonmusician” groups; and exploration of which specific musical skills (rhythm, pitch perception, timbral discrimination) most strongly relate to speech perception under different kinds of degradation.
Source: Julia Majors – Acoustical Society of America
Publisher: Organized by NeuroscienceNews.com.
Image Source: NeuroscienceNews.com image is in the public domain.
Original Research: The study was presented at the 174th meeting of the Acoustical Society of America.
Acoustical Society of America, “Is There a Musical Method For Interpreting Speech?,” NeuroscienceNews, 9 December 2017. Retrieved December 9, 2017.