Abstract

This article asks what the emerging computational field of sentiment analysis can teach us about the sentimental novel, and vice versa. It argues that, despite humanistic skepticism about quantitative methods and sentiment analysis’s well-known limitations (in recognizing irony, for example), sentiment analysis can help us better to understand the novel form and the sentimental novel in particular. The literary approach to computational analysis taken in this article demonstrates the ability of sentiment analysis to link large-scale observations about text data to small-scale features of individual texts and reveals that the sentimental novel itself already constitutes an analytical tradition.