A listener's explicit knowledge of the links between phonetic variants and social categories can be referred to as their sociophonetic knowledge. The goal of this dissertation was to directly examine the degree to which sociophonetic knowledge influences listeners' ability to process and encode lexical tokens produced in different accents, with the aim of informing cognitive models of speech processing. Experiments 1a and 1b controlled listeners' sociophonetic knowledge by exposing them to a novel accent of English and varying the social categories they were taught to associate with the accent. In a speeded auditory lexical decision task (Experiment 1a), the social information manipulation was not found to influence response patterns; therefore, this experiment failed to provide evidence that sociophonetic knowledge impacts the speed and accuracy of lexical processing. However, in an explicit lexical recognition memory task (Experiment 1b), response times were slower for participants who were given low-status social information compared with those who were given high-status social information or no information, demonstrating that the specific sociophonetic knowledge that listeners hold about an accent can impact the robustness with which they encode lexical tokens in memory. The differences in the results between the two tasks may be due to task-specific processing demands. In Experiment 2, listeners were exposed to stimuli produced in the Midland, Northern, and Southern varieties of American English, which are subject to different degrees of sociophonetic knowledge across the general population. During an explicit lexical recognition memory task, some participants were told the regional backgrounds of the talkers while others were not. It was predicted that encoding would be strongest for tokens produced in the accent recognized as ideologically standard, Midland English, but also that being given regional information about the talkers would increase the robustness of encoding, particularly for nonstandard varieties about which listeners held sociophonetic knowledge. Analysis of participants' response times provided evidence of a standard dialect benefit across listeners from different regions, but performance for some listeners was unexpectedly hindered by the presence of social information about the talkers. The results indicate that social information is highly integrated with speech processing, but that it is not facilitative in all situations, even for varieties that are subject to sociophonetic knowledge. Together, these results demonstrate that differences in sociophonetic knowledge have the ability to impact the robustness with which tokens are encoded in memory. These results overall are consistent with exemplar theoretic models of speech processing in which the social status attributed to different phonetic variants influences the relative strength of memory traces. At the same time, the results of these experiments also indicate sociophonetic knowledge does not facilitate speech processing in all instances. [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.]