A Linguistic Analysis on Negation and Emotion Shift

Emotion classification is gaining more attention in the field of linguistics, to
facilitate the development of automatic emotion analysis and classification. While many focus on how emotions can be better classified, there are also other linguistic devices that contribute to emotions, for example negation. This paper sees negation as an important linguistic device that causes shifts of emotions and highlights the importance of taking negation into account during the automatic emotion analysis process. The paper takes data from the Chinese online social media platform Sina Weibo as the corpus, then observes the interaction between the two major negators bu and mei and their modifying emotion expression. The finding reveals that negators have a significant contribution to emotion shifts in social media discourse, and that the type of emotion shifts varies due to different morphological features, semantic differences or pragmatic uses such as irony.