Reading Human Nature: Literary Darwinism in Theory and Practice

9. Quantifying Agonistic Structure in The Mayor of Casterbridge

177 Models of Tragedy in the Interpretive History of Mayor We set up a website for The Mayor of Casterbridge separate from the larger website listing about two thousand characters from about two hundred novels. To distinguish the two sites, we referred to the larger website as the “multi-novel website.” Our aim in setting up an individual site for Mayor was to collect data on enough characters from a single novel to give a comprehensive analysis of the organization of characters and reader responses in that one novel. We chose Mayor in part because it is relatively compact, has only a few major characters, and has characters who are very distinctively marked in motives and personality. We listed six characters from Mayor: Henchard (the title character); his wife Susan; his stepdaughter Elizabeth-Jane; his rival Donald Farfrae; Lucetta, the woman for whose favors Henchard and Farfrae enter into competition ; and Newson, the sailor who, at the beginning of the novel, buys Henchard’s wife and daughter from him. Another reason for selecting Mayor as a case study is that it has an unusual agonistic and tonal structure. By using the average scores of the multi-novel website as a frame of reference, we hoped to tease out the structural peculiarities of Mayor and draw illuminating interpretive inferences from those peculiarities. chapter 9 , Quantifying Agonistic Structure in The Mayor of Casterbridge Joseph Carroll, John Johnson, Jonathan Gottschall, Daniel Kruger, and Stelios Georgiades 178 Reading Human Nature Interpretive commentary, and especially the interpretation of tone, is often regarded as a form of study too subjective and impressionistic ever to be brought within the range of quantification and empirical analysis. By giving a quantitative analysis of the tone in a single novel, we aimed to demonstrate that there need be no aspect of literary study inaccessible to empirical study, and further, that quantification could confirm, refine, correct, and develop the insights of traditional interpretive criticism. We solicited participation in the Mayor study by directly contacting scholars who had published on Hardy and particularly on Mayor or on other Hardy novels. We also advertised the study on the listserv of the Thomas Hardy Association and listservs associated with the study of Victorian literature. All participation was anonymous, but we collected information about respondents’ age, sex, level of education, when and why they read the novel, and whether they had published on Mayor or other works of Hardy. By analyzing this information, we determined that a total of eighty-five individual coders responded to the survey. Fifty-one were males, thirty-four females. The youngest respondent was twenty-three, and only eight respondents were under the age of thirty. All had college degrees. Nine had a bachelor’s degree, twenty-one a master’s, and fifty-five a doctorate. Twenty-five had published on Mayor; another twenty-three had published on some other novel by Hardy; and another ten had published on some other aspect of Hardy’s work. Thus, a total of fifty-eight out of the eighty-five (68%) had published on some aspect of Hardy’s work. Sixty-seven respondents reported having read the novel within the past five years, and thirty-one within the past year. Fifty-five read it either for teaching a class or for “professional purposes.” In sum, almost all the respondents were very familiar with the novel. A number of respondents completed more than one protocol, and a total of 124 protocols were completed. To assess the level at which respondents agreed in their assessments of the characters, we conducted “alpha reliability estimates .” In most psychological research, alpha values around .70 are considered acceptable, and alphas in the .80 to .90 range are considered good. Values above .90 are normally achieved only by trained professionals. The average alphas across all categories for the Mayor respondents is .84. The lowest alpha values were for a minor character (Newson), who received only five codings. If we Agonistic Structure in The Mayor of Casterbridge 179 exclude Newson’s alpha values, the average alpha values across all categories is .89. In other words, there was a high level of consensus among the respondents on all the substantive categories of analysis. Agonistic role assignments are a different matter. We discuss those at the beginning of the next section. To orient readers who have not read Mayor or have not read it recently, we shall concisely summarize the plot. The...