Methods and Evidence in Psychonarratology and the Theory of the Narrator: Reply to Diengott

Abstract

Methods and Evidence in Psychonarratology and the Theory of the Narrator:Reply to Diengott Peter Dixon (bio) and Marisa Bortolussi (bio) We welcome this opportunity to reply to Nilli Diengott's review of our book, Psychonarratology: Foundations for the Empirical Study of Literary Response. While she has identified a number of regrettable misstatements in the book and has raised a number of interesting arguments concerning our hypotheses about readers' representations, it also appears that there are a number of aspects of our methodology and conceptual analysis that may not have been understood. We hope that the present reply will promote a better understanding of our approach in order to foster productive debate. Although Diengott's review concentrates on our treatment of the narrator, according to our analysis of her comments, there are really three levels of concerns. First and most importantly, there are issues pertaining to methodology, primarily having to do with the distinction between textual features and reader constructions and how these concepts manifest themselves in empirical research. Second, there are objections to our hypotheses concerning the reader's representation and processing of the narrator. These are crucial concerns since our analysis of the narrator is fundamental to the theoretical developments throughout the book. Finally, there are questions concerning the analysis of "transparency" and its relationship to identification. In what follows, we discuss each of these concerns in turn. We then comment on a few other, more miscellaneous points raised in the review. [End Page 317] Methodology Perhaps the single most fundamental element in our approach to the empirical study of literary response is the distinction between that which is in the text (features) and that which is in the reader's head (constructions). Features are characterizations of the text that, at least in principle, could be objectively defined and that would be true of a given text regardless of who analyzes the text and the context in which that analysis occurs. Constructions, on the other hand, are subjective and may vary with the reader, her knowledge and experience, and her current goals and circumstances. This distinction is crucial because it allows one to match method to question. Features are a matter of definition and, as a consequence, are more or less useful, well-defined or not so well-defined, precise or imprecise, and so on. One can take poor features (e.g., those that are imprecise or poorly defined) and hone them via argument, analysis, and elaboration. However, they cannot be dismissed as incorrect; they are merely a matter of definition, after all, and any number of features may be defined even if they are not particularly helpful. In contrast, reader constructions are potentially in the mind of the reader, and it is a matter of empirical evidence as to whether and under what circumstances they occur. Various kinds of evidence might be marshaled for a reader construction, and such evidence might be strong or weak or more or less convincing. However, ultimately the status of a reader construction is likely to rest on the results of carefully designed studies rather than on logical argument. As an example of this distinction, consider the narratological concept of the implied author. As defined by Gerald Prince, paraphrasing Wayne Booth, the implied author is "the author's second self, mask, or persona as reconstructed from the text" (000). The problem with this definition (and, in our view, with many other narratological concepts) is that it is not clear whether the implied author should be thought of as a textual feature or as a reader construction. A word such as "reconstructed" would seem to allow for subjective inference processes on the part of the reader, implying that the term corresponds to a reader construction; however, it is also possible to envision that a logical, externally validated analysis of the text could be used to identify aspects of the implied author. For example, Iser understood the implied author in interactive terms, so that it emerged on the basis of the interaction of the text and the reader. However, Iser was never concerned with the real reader, and his conception was idealized in the sense that it attempted to provide an account for the...