Nonnative speech is a naturalistic source of speech variation that is relatively familiar to listeners while posing significant perceptual challenges. This dissertation examined how listeners represent and process nonnative speech. Study 1 investigated the salient perceptual dimensions of nonnative speech for native listeners using a free classification task. Accent strength was a highly salient feature of nonnative speech across multiple stimulus sets. Accent strength remained a salient organizational schema even when listeners were instructed to classify talkers by native language background. To examine the effects of familiarity with different accents on nonnative speech perception, Study 2 tested native English, Korean, and Spanish listeners' classification of native and nonnative talkers. These listener groups performed similarly in overall classification accuracy and demonstrated heightened perceptual sensitivity to talkers with whom they shared a native language. An accent discrimination task was employed in Study 3 to examine the process by which listeners differentiate one accent from another. Native listeners were sensitive to the differences between two nonnative accents, albeit to a significantly lesser degree than to the difference between native and nonnative accents. Sensitivity levels also differed across accent pairs and stimulus types. Finally, the relationship between listeners' accent representation and accented-speech understanding was assessed in Study 4, using accent categorization and nonnative speech-in-noise transcription tasks. Listeners' accent categorization performance predicted their transcription performance. Listeners' foreign language study experiences predicted their accent categorization performance, but not their transcription performance. More variable linguistic experiences, such as those gained through foreign language study, may lead listeners to have better-specified nonnative accent categories; however, those same experiences do not improve their accented speech understanding. Together, these studies suggest that listeners' perception of nonnative accents is organized along a graded scale of perceptual similarity to the listeners' own accent, as well as to the accent in their environment. Listeners' linguistic environments and experiences, therefore, are central to shaping their representations of nonnative accents. |