Linguistics and literature inform one another, since every writer is a user of language, with a more or less developed perception of his/her particular grammar, and every written work interacts with the linguistic capacities and prejudices of its audience (as well as itself serving as an object of linguistic analysis and a potential model for future readers). I have conducted extensive research on a specific linguistic phenomenon, the instrumental singular feminine variation between –oju (–eju) and –oj (–ej) in nineteenth–century Russian belletristic prose. In this paper, I discuss Tolstoy's contrastive use of the two instrumental variants of the word
A sample of all –a nouns in the first two books of Anna Karenina shows that Tolstoy uses the more archaic –oju (–eju) only 12% of the time. Yet in this same sample, zhenoju, though not the dominant form, occurs 9 of 27 times, or 33% of the time, while the other –a nouns which occur with any regularity appear far less frequently in their older variants: golovoju (1 of 18, or 6%), rukoju (1 of 12, or 4%), ulybkoju (2 of 61, or 3%). These figures indicate that the difference in forms may be employed for some literary effect. Some might argue that the word itself may have had a particular tendency to retain the more archaic ending. While such might be the case, the centrality of the concept of wife to the novel, the evidence of other literary works from the period (in which the relative frequency of zhenoju is far lower compared to that of zhenoj), and the existence of locutions in which both variants occur in the same context in the novel point toward a conscious playing on the variation for narrative effect. (That Tolstoy was acutely aware of the existence of the variation and the author's role in choosing forms can be seen from the changes in his use of the same variant morphology in instrumental feminine adjectives: in his early works, the archaic longer forms in –oju (–eju) account for only 19% of all forms, in the big novels of the 1860s and 1870s these archaic forms dominate (94%), and in his later works they all but disappear.)
In specific, the change between zhenoj and zhenoju denotes a subtle shift of focus from the intimate, private sphere to the formal, public sphere. It can be seen in the mental processes of Stepan Arkad'ich and Dolly, where zhenoj predominates and zhenoju indicates that the relationship is being seen in some sense from outside; in Levin's musings on marriage with Kitty, where again zhenoj predominates; and in the thoughts of Aleksej Aleksandrovich about Anna, where zhenoju predominates and zhenoj indicates the moments in which he allows himself to think about intimacy rather than public perception. The body of the paper presents examples of these shifts and argues for the significance of authorial awareness of linguistic signs as narrative devices which will be intelligible to the audience on some level, however subliminal.