The core objective of this study was to examine whether acoustic variability of vowel production in American English, across speaking tasks, is systematic. Ten male speakers who spoke a relatively homogeneous Wisconsin dialect produced eight monophthong vowels (in hVd and CVC contexts) in four speaking tasks, including clear-speech, citation form, passage reading, and conversational speech. Motivated by the phenomenon of vowel reduction and the Hyper- and Hypo-speech theory of speech production, the hVd context in a clear-speech task was operationally defined as the reference point for the evaluation of reduction effects. The first three vowel formant frequencies (F1, F2, F3) measured at the temporal midpoint of English monophthongs, and the derived Euclidean distances on the F1-F2 and F2-F3 planes, demonstrated an overall centralization effect of the vowel spaces across speaking tasks, with the largest vowel space generally observed in clear-speech, followed by citation, then reading, with the smallest vowel space produced in conversational speech. While the coarse-grained, cross-task variation was similar overall for all speakers, substantial fine-grained individual differences were found in the patterns of formant frequency changes. These findings were discussed within the framework of reduction and H&H theory, and relevant issues concerning potential implications for speech motor control were addressed. An exploratory examination of five diphthongs (/ei/, /alpha I/, /alpha upsilon/, /oupsilon/, and /[open o]i/) in American English produced by two male speakers (from the speaker pool of the vowel study) was also incorporated into this investigation as a preliminary evaluation of diphthong variability. The diphthongs were produced in the same consonant contexts and speaking tasks as the monophthong vowels. Motivated by the concept of diphthongs as unique gestures, diphthong F2 transitions were examined for acoustic characteristics consistent with gestural reduction. Relatively systematic changes in line with the prediction of reduction were found for diphthongs /ei/ and /alpha I/, but not the remaining three diphthongs /alpha upsilon/, /oupsilon/, and /[open o]i/. The effectiveness of a gesture model in characterizing diphthong changes was considered, and directions for future research were proposed. [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://www.proquest.com/en-US/products/dissertations/individuals.shtml.]