AntConc와 Python SnowNLP를 활용한 ‘香’의 인지⋅화용적 분석 — ≪三國演義≫, ≪水滸傳≫을 중심으로

Abstract

This paper analyzes the collocational structure of the character ‘香’(xiang, fragrance) and its associated emotional aspects as depicted in two representative novels of the Ming(明) periods, Romance of the Three Kingdoms (≪三國演義≫) and Water Margin (≪水滸傳≫). The purpose of the study is to elucidate the cognitive mechanisms underlying semantic expansion derived from collocational relationships involving ‘香’, and to examine the emotional implications embedded in this process from a pragmatic perspective. To achieve this aim, collocational structures were extracted using AntConc, and quantitative sentiment analysis was conducted using Python SnowNLP. The analysis revealed that in Romance of the Three Kingdoms, the character ‘香’ frequently occurs in strategic and tense contexts, often evoking negative or tense emotions, yet in certain instances, it exhibits a broader range of emotional positivity. In contrast, Water Margin predominantly associates ‘香’ with negative emotions, functioning primarily as a device for expressing social conflicts and tension. This research contributes by integrating digital humanities tools for sentiment analysis with cognitive linguistics and pragmatics, thereby elucidating the conceptual emotional frameworks embedded within Ming literary texts.