Limits to Interpretation: The Meanings of Anna Karenina

6. From Theory to Practice

63 6. From Theory to Practice Before turning to the opening of Anna Karenina, I would like to clarify the relation between the methodology I advocate and the results I claim for it by explaining briefly the procedure I followed to produce the map of possible readings (see also 5.3). A seamless transition from theory to praxis probably requires more self-consciousness and self-control than most investigators are likely to possess; and much theory is based on abstraction that leaves out some of the fine-grained specifics that make individual works of literature what they are. It is to be expected, therefore, that at least some dehiscence will occur between any methodology and its results. Nevertheless, awareness of this problem may help to mitigate it, especially in the case of a methodology that is inherently adaptive and that is based on linguistic phenomena inherent in discourse. I started by reading through the novel several times and marking in the margins all of the textual moments that fit the criteria of hermeneutic indices (5, 5.1, 5.2); these numbered roughly 1,600. I transferred references to these moments onto index cards, noted their implications for the novel’s array of meanings, and then sorted the cards by the categories that appeared to emerge from the novel itself, that is, the narrator, characters, structure, plots, themes, scenes, recurring imagery, and so on. Finally, I arranged these categories into a sequence that begins with the implications of the novel’s opening lines and pages, which are especially important because they shape the reader’s expectations with regard to the remainder of the text, and then goes on to the novel’s portrayal of reading, perceiving art, and construing meaning in general, all of which have a bearing on the reader’s attempt to understand the novel in its own terms. The length of my discussions of various other topics tends to be a function of the extent to which the novel emphasizes them. Thus, I analyze the relations among characters in some detail because this is what the novel is mostly about. At all stages I adjusted the meanings of individual indices when new or different connections among them seemed to warrant it. My procedure thus flows naturally from my methodological commitment to suppressing as much as possible all a priori projections onto the novel and to grasping how meanings emerge in it from the bottom up. This procedure is actually the only way to deal with a large amount of variegated data of any kind if one believes in trying to understand its defining characteristics. My focus on the textual moments I call “hermeneutic indices” also grows out of the general task I set myself. As my examples above show (5.1), hermeneutic indices are especially telling instances of meaning creation because they reveal something about the frames of reference that underlie the text (which obviously implies that they are not all the instances of meaning creation in a text). They are among the most important elementary units on the most basic textual level out of which the higher, more complex textual meanings are composed, and their importance is a function of the larger quantity of information they reveal by virtue of pointing beyond themselves to the semantic fields they imply (via the juxtaposition or translation that defines them). Thus, as we will see, the equation in Anna Karenina’s first line between family happiness and resemblance is far more revealing of some of the novel’s larger meanings than is the characterization of Stiva shortly thereafter as having a body that is full and well tended, because the first involves the unexpected juxtaposition of two semantic fields, while the second does not. But despite the explanatory power of hermeneutic indices, it is important to remember that they are merely variants of Jakobson’s “metalingual function” in the sense of being comparable to the overt kinds of recoding or glossing that he identifies (4). Consequently, by focusing on hermeneutic indices I do not claim that I have discovered an unknown feature of language ; instead, I advocate a systematic exploitation of a feature of language that has been neglected in textual interpretation. Moreover, given that glosses are inherent in utterances, we cannot expect hermeneutic indices always to stand out in the text as if they were either heterogeneous elements in it or potential knots of evidence waiting for some belief to illuminate their true nature...