This thesis describes a generative model for representing disfluent phenomena in human speech. This model makes use of observed syntactic structure present in disfluent speech, and uses a right-corner transform on syntax trees to model this structure in a very natural way. Specifically, the phenomenon of speech repair is modeled by explicitly modeling the partially completed constituent structure, building up phrasal structure as it is read in, and allowing incomplete constituents as part of the final structure of an utterance. This model is then implemented in a working system for parsing transcribed text. This system maps transformed syntax trees to a hierarchical time-series model, which is able to parse in linear time by implementing a fixed-depth stack. The results of this system dramatically improve parsing and repair detection over standard baselines, and are approaching the performance of models that are not constrained by incremental processing and psycholinguistic plausibility. These results are then analyzed, and the parameters to this parser are varied systematically to examine the effect of various settings and models of repair. Several extensions to the model are introduced which incorporate additional features such as lexical identity while remaining incremental, suggesting suitability for extended models of disfluency as well as providing a starting point for psycholinguistic models of disfluency. Finally, this thesis concludes by laying out several possible future research directions, and by summarizing the work produced so far. [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.]