Vowel nasalization usually occurs through a two-step process whereby a vowel is nasalized via coarticulation with a nearby nasal segment; when the language later drops the nasal segment, a nasal vowel remains. Spontaneous vowel nasalization is a rare, peculiar form of nasalization that emerges in contexts that lack an historical etymological nasal (Blevins and Garrett, 1992; Matisoff, 1975; Ohala, 1975, 1974). Spontaneous vowel nasalization has been found to emerge in contexts with glottal consonants, low vowels, and transitionally breathy segments such as fricatives and aspirates. Blevins and Garrett (1992) classified these contexts as rhinoglottophilia 'nose-larynx-affinity', rhinochthamalophilia 'nose-low/flat-affinity', and rhinosyrigmatophilia 'nose-whistle-affinity'. Many articulatory and acoustic explanations have been proposed for spontaneous nasalization. First, vowels may nasalize near glottal consonants because they are underspecified for velopharyngeal opening and thus may be produced with a lowered velum. Second, low vowels may nasalize because they can better tolerate the acoustic consequences of greater nasal coupling; the intraoral acoustic impedance of low vowels is naturally lower. Finally, speakers may nasalize vowels in breathy environments through misperception or enhancement because nasality and breathiness are acoustically similar (Matisoff, 1975; Ohala and Amador, 1981). Due to the acoustic similarities between nasalization and breathiness, e.g. high spectral tilt and weakened F1, disentangling the relative roles of physiological nasalization and breathiness during spontaneous nasalization is not possible using acoustic data alone. There have been few articulatory investigations that assess the origins of spontaneous nasalization (Clumeck, 1975; Ohala, 1971). Thai arguably manifests an interaction among rhinoglottophilia, rhinochthamalophilia, and rhinosyrigmatophilia-based spontaneous nasalization. Thai is known to nasalize low and mid-low vowels after the voiceless glottal fricative /h/ and glottal stop /?/, such that /h[epsilon]-l/ [h[epsilon]-l] 'parade' and /?aw?/ [?aw-l] 'to take' are reported to sound nasal (Cooke, 1989; Matisoff, 1975; Noss, 1964). Vowels following /h/ are reportedly more susceptible to nasalization than those following /?/ (Matisoff, 1975; Cooke, 1989). Furthermore, of the nasalized vowels, Cooke (1989) observed that /a/ is reported to sound more nasal than /[epsilon]/ and /[open-mid back rounded vowel]/. In Thai it is possible that the velopharyngeal underspecification of glottal consonants, low tongue height, and breathy-nasal acoustic similarity (in proximity to /h/) may all play a role in nasalization. In order to deepen our understanding of these potentially interacting sources of spontaneous nasalization in Thai, this dissertation undertakes an extensive articulatory investigation of physiological nasalization and breathiness during spontaneously nasalized vowels in Central Thai. Spontaneous nasalization has been an historically difficult topic of study because its analysis requires a combination of articulatory data that measures both nasalization and phonation quality. This dissertation integrates measures from state-of-the-art ultra-fast magnetic resonance imaging (MRI), aerodynamics, and electroglottography (EGG) to estimate physiological nasalization and breathiness. The speech of ten speakers of Central Thai was assessed using these tools: Four speakers were recorded using ultra-fast MRI and six speakers were recorded using aerodynamics and EGG. Speakers produced CV syllables that varied by onset consonant and vowel height within the same carrier phrase. Syllables with an onset glottal consonant /h, ?/ were varied at four different vowel heights; syllables with onset consonants /n, th, d/ were included as nasal and oral controls. The major finding of this dissertation is that /h/-onset syllables are produced with greater physiological nasalization than /?/-onset syllables. This finding suggests that, counter to previous claims that glottal consonants are generally underspecified for velopharyngeal opening in Thai, /?/ is in fact specified, at least in a prosodically prominent context. Furthermore, using EGG we observed a slight increase in breathiness during vowels just after /h/, but not after /?/, suggesting a potential relationship between nasalization and breathiness in Thai. The voiceless turbulence of /h/ induces coarticulatory breathiness during the following vowel; this breathiness may facilitate the perception of nasalization during the following vowel. This might explain why vowels after /h/ reportedly sound more nasal than vowels after /?/ (Cooke, 1989; Matisoff, 1975). Further perceptual testing is needed to assess this possibility. Furthermore, we observed inconsistent and often minimal variation in physiological nasalization during vowels of varying height in both nasalized and non-nasal context. This suggests that spontaneous vowel nasalization in Central Thai may now be similar to Northeastern Thai, a dialect where all vowels were reported to nasalize after glottal consonants approximately 40 years ago (Matisoff, 1975). Finally, during /hV/ syllables, all /h/ consonants are produced with greater nasal airflow than the following vowel. The onset /h/ appears to be the locus of nasalization that spreads to the vowel through coarticulation. The results of this dissertation deepen our understanding of the production of spontaneous nasalization and contexts that facilitate its realization in Thai and perhaps cross-linguistically as well. Our results demonstrate that spontaneous nasalization in Thai is primarily attributed to rhinoglottophilia: velopharyngeal underspecification of onset /h/. Rhinosyrigmatophilia may potentially also play a role: the presence of breathiness after /h/ may enhance the percept of nasalization. Nasal coupling and breathiness may be integrated into a single acoustic object that the listener perceives as nasal. This possibility presents an intriguing example of the many-to-one problem, whereby many possible articulatory configurations may result in similar acoustic output (Maeda, 1990). [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]