Background: Auditory-perceptual assessment of voice is a subjective procedure. Artificial intelligence with deep learning (DL) may improve the consistency and accessibility of this task. It is unclear how a DL model performs on different acoustic features. Aims: To develop a generalizable DL framework for identifying dysphonia using a…

Variability is intrinsic to human speech production. One approach to understand variability in speech is to decompose it into task-irrelevant ("good") and task-relevant ("bad") parts with respect to speech tasks. Based on the uncontrolled manifold (UCM) approach, this dissertation investigates how vowel token-to-token…

Purpose: Vocal roughness is often present in many voice disorders but the assessment of roughness mainly depends on the subjective auditory-perceptual evaluation and lacks acoustic correlates. This study aimed to apply the concept of roughness in general sound quality perception to vocal roughness assessment and to characterize the relationship…

This dissertation analyzes the production and perception of lexical stress in trilinguals' first, second, and third languages (L1, L2, and L3) to evaluate how the cue-weighting transfer hypothesis applies to L3 acquisition. According to this hypothesis, acoustic cues to stress have different weights across languages, and results from both…

Purpose: The speech motor system uses feedforward and feedback control mechanisms that are both reliant on prediction errors. Here, we developed a state-space model to estimate the error sensitivity of the control systems. We examined (a) whether the model accounts for the error sensitivity of the control systems and (b) whether the two systems…

This dissertation investigates language-specific acoustic and aerodynamic phenomena in language contact situations. Whereas most work on second language and bilingual phonology has focused on individual consonants and vowels, this project examines patterns of coarticulation in the two languages of Spanish-English and French-English bilingual…

This dissertation explores a purportedly unusual word type known as the "sesquisyllable," which has long been considered characteristic of mainland Southeast Asian languages. Sesquisyllables are traditionally defined as "one and a half" syllables, or as one "major" syllable preceded by one "minor" syllable,…

The present study reports on the findings of a cross-sectional acoustic study of the production of Spanish vowels by three different groups of speakers: 1) native Spanish speakers; 2) native English intermediate learners of Spanish; and 3) native English advanced learners of Spanish. In particular, we examined the production of the five Spanish…

Purpose: With the goal of using articulatory adjustments to reduce hypernasality, this study utilized an articulatory synthesis model (Childers, 2000) to simulate the adjustment of articulatory configurations with an open velopharynx to achieve the same acoustic goal as normal speech simulated with a closed velopharynx. Method: To examine the…

In an attempt to understand what acoustic/auditory feature sets motivated transcribers towards certain labeling decisions, I built machine learning models that were capable of discriminating between canonical and non-canonical vowels excised from the Buckeye Corpus. Specifically, I wanted to model when the dictionary form and the transcribed-form…

Speech production is inextricably linked to speech perception, yet they are usually investigated in isolation. In this study, we employed a verbal-repetition task to identify the neural substrates of speech processing with two ends active simultaneously using functional MRI. Subjects verbally repeated auditory stimuli containing an ambiguous vowel…

The goal of this dissertation is to introduce a method for deriving morphemes from Arabic words using stem patterns, a feature of Arabic morphology. The motivations are three-fold: modeling with morphemes rather than words should help address the out-of-vocabulary problem; working with stem patterns should prove to be a cross-dialectally valid…

Lexical-tone languages use fundamental frequency (F0/pitch) to convey word meaning. About 41.8% of the world's languages use lexical tone (Maddieson, 2008), yet those systems are under-studied. I aim to increase our understanding of speech-sound inventory organization by extending to tone-systems a model of vowel-system organization, the Theory of…

A variety of studies have demonstrated that organizing stimuli into categories can affect the way the stimuli are perceived. We explore the influence of categories on perception through one such phenomenon, the perceptual magnet effect, in which discriminability between vowels is reduced near prototypical vowel sounds. We present a Bayesian model…

The [voice] distinction between homorganic stops and fricatives is made by a number of acoustic correlates including voicing, segment duration, and preceding vowel duration. The present work looks at [voice] from a number of multidimensional perspectives. This dissertation's focus is a corpus study of the phonetic realization of [voice] in two…