This dissertation looks at two works of nineteenth-century British fiction that are considered outliers: Sir Walter Scott's Saint Ronan's Well (1824) and George Eliot's "The Lifted Veil" (1859). Saint Ronan's Well, a work of domestic fiction, has long been described as unusual for Scott because it is unlike his signature historical novels. "The Lifted Veil,” a Gothic novella, is generally understood as different in kind from Eliot’s realist and social problem fiction. I describe both texts as outliers because they have been described as atypical (in the case of Eliot) or less worthy of study (in the case of Scott) by scholars, for myriad reasons. My work uses both computational methods and tools and traditional literary close readings to test and assess these outlier works. I use stylometry, a computational tool that reads and compares texts to determine authorship attribution, to determine if both texts are indeed outliers for these authors. In addition, I use stylometric methods to analyze claims made by initial reviewers and contemporary critics about comparative authors and genres for Saint Ronan's Well and "The Lifted Veil." I examine statistical or stylistic evidence to test whether those long-standing claims of literary difference are supported with computational evidence. Each chapter consists of a series of stylometric tests that are analyzed in conjunction with close readings of text. The dissertation reaches three conclusions, based on the results of stylometric tests described across its four chapters. First, I find that, although Saint Ronan's Well is written in a unique subgenre for Scott, it is statistically and stylistically similar to his other novels. Second, I argue that "The Lifted Veil" is both an outlier for Eliot and an outlier among canonical work of the period in general, as indicated by the results of several stylometric tests. Finally, I argue that focusing on literary outliers is a necessary and productive step forward for traditional and computational literary studies. Focusing on texts that are literary and statistical outliers shows how computational and traditional literary methods can blend together to test the extent of generalizable knowledge in literary studies, especially with nineteenth-century British fiction.
Included in this item (2)
- Partial requirement for: Ph.D., Arizona State University, 2021
- Field of study: English