Emotion plays a central role in understanding Joyce's Dubliners and remains the focus of scholarly debate. Adopting a corpus stylistic approach, this study uses self-built corpora and the Wmatrix semantic analysis tool to examine the foregrounded features of emotive language in this short story collection from the dimensions of both external and internal deviation. The study reveals that, in terms of external deviation, emotion density in Dubliners is generally lower than in the reference corpus, while the frequencies of two positive emotion types, happiness and liking, are notably higher. The study also reveals that, in terms of internal deviation, emotion density varies significantly across individual short stories, along with noticeable differences in the frequency and diversity of emotion types. The study will shed light on the complex texture of emotive language in Dubliners while introducing a new method and perspective for studying emotion in literature.

Key words: corpus stylistics; foregrounding; emotion density; emotion type; Dubliners