What reduced speech teaches us about speech processing
Mirjam Ernestus, Centre for Language Studies, Radboud University Nijmegen, the Netherlands
Abstract: In casual conversations, words are often produced with weakened segments or with fewer syllables than in formal speech. For instance, in everyday speech, the words yesterday is often pronounced like yeshay. I will first introduce the phenomenon of reduced speech. I will then shortly discuss why reduced speech is problematic for automatic speech recognition. Finally, I will discuss a series of corpus and experimental studies that show a correlation between reduction, on the one hand, and the speaker’s and listener’s cognitive load during conversations, on the other.
———————————————————————————————————————————————
The role of rate and rhythm in speech perception
Hans Rutger Bosker, Max Planck institute for Psycholinguistics, Nijmegen, the Netherlands
Abstract: Speech comprehension is an intriguing type of human behavior, not in the least because of its resilience to noise. I will describe a neurobiologically plausible model of how the temporal properties of the speech signal help the listening brain overcome intrinsic and extrinsic sources of noise. This model points at a central role for neural oscillations, phase-locking to the temporal modulations in the speech signal (neural ‘speech tracking’), and contributing to the initial parsing and decoding of discrete linguistic information from the continuous speech stream. I will draw from acoustic, psychoacoustic, and neuroimaging data from experiments investigating (1) how listeners comprehend speech produced at various speaking rates (i.e., signal-intrinsic noise); and (2) how listeners comprehend speech-in-noise (i.e., signal-extrinsic noise).
———————————————————————————————————————————————
The pupil dilation response to auditory stimuli: Current state of knowledge
Adriana A. Zekveld, Thomas Koelewijn, Sophia E. Kramer, Section Ear & Hearing, Dept. of Otolaryngology-Head and Neck Surgery, Amsterdam Public Health research institute, VU University Medical Center, the Netherlands
Hearing impairment can result in difficulties with understanding speech in everyday listening situations, especially in the presence of background noise. To compensate for speech recognition difficulties, cognitive abilities like attention, working memory and linguistic processing are relevant. Although the use of these functions facilitates speech perception, it can result in increased listening effort. The need to put additional listening effort in all kinds of listening situations can eventually lead to fatigue. Measuring changes in pupil diameter (pupillometry) is a physiological method of assessing listening effort. The pupil size increases when listening effort increases and thereby indicates how much effort a particular listening task costs. In recent years, the number of studies inside and outside the field of hearing science addressing the pupil response evoked by auditory stimuli has increased considerably. We will present the results of a review that aimed to provide an exhaustive overview of the existing research describing the pupil response to auditory stimuli.
Results: In total, 141 studies were included and discussed in this review. These originated from multiple domains, including hearing science and linguistics. Furthermore, the review covers pupillometry research into motivation, memory, and emotional processing. We will present a unique overview of these studies, organized according to the components of the Framework of Understanding Effortful Listening (FUEL). The results show that the pupil response is sensitive to various task manipulations (e.g. intelligibility level, linguistic complexity, emotional valence) as well as interindividual differences such as age and hearing status. Many of the findings have been replicated. Frequent interactions between the independent factors affecting the pupil response (like degradation level, sample characteristics, reward level, event probability) have been reported, which suggests that multiple processes compete for resources when listening to sound. The interplay between these processes should be taken into account in future studies. Also, research could focus more on interindividual differences, for example by including middle-aged and older participants. The present review facilitates careful design of future studies by indicating what factors should be controlled for. Besides presenting the main findings of the literature, we will also highlight several studies performed inside and outside the field of hearing science that have interesting implications for the application of pupillometry in the assessment of listening effort.
In conclusion, measuring the pupil dilation response to auditory stimuli has been demonstrated to be a sensitive method applicable to numerous research questions. The sensitivity of the measure calls for carefully designed stimuli.