The Time-Course of Linguistic Interference on Measures of Speech-in-Noise Perception and Listening Effort

Speech perception is a dynamic process whereby listeners must attune to a target talker. In adverse listening conditions, listeners must overcome additional challenges associated with energetic and informational masking from competing sound sources. Linguistic interference, a type of informational masking resulting from the linguistic content of competing speech, is the main topic of this thesis. Many studies of linguistic interference assess how different types of competing speech impact speech perception, without considering a listener’s ability to adapt to adverse listening conditions over time. This thesis investigates the change in linguistic interference over time in monolingual and multilingual listeners. The first four experiments investigate how knowledge of the linguistic content of competing speech impacts performance and how the patterns of linguistic interference depend on whether listening takes place in native and non-native languages. The last three experiments look specifically at monolingual listeners and investigate not only how speech-in-noise performance changes over time, but also how physiological and subjective measures of listening effort and fatigue change as listeners adapt to adverse listening conditions. Pupillometric measures of listening effort are used to detect changes in physiological state potentially outside of conscious awareness. Effort is measured while listening to a talker in intelligible and unintelligible competing speech and are compared to speech-in-noise performance, and subjective self-reports of listening effort and fatigue. The results from these experiments demonstrate general listener abilities to improve in speech-in-noise perception, and most experiments demonstrate similar rates of improvement over time for both intelligible and unintelligible speech maskers. Task-evoked pupil responses also generally decreased over time. These experiments demonstrate how speech intelligibility, SNR and experimental designs can significantly affect pupillometric measures, and how they can inform empirical and theoretical considerations when measuring effort, fatigue and motivation in speech-in-noise perception.