A Computational Linguistics Framework for Quantifying Affective Fidelity in Classical Literary Translation: A Case Study on English Translations of Fusheng Liuji

The evaluation of affective fidelity in literary translation has long relied on subjective assessment, lacking scalable and reproducible quantitative methods. To address this gap, this paper proposes and implements a novel computational linguistic framework designed to quantify affective fidelity in English translations of classical literature. The framework constructs standardized multi-dimensional affective vectors for each translation by integrating multiple lexicon-based sentiment analysis tools (e.g., VADER and TextBlob). Building on this, two core metrics are introduced: first, the Affective Consensus Score (ACS), which measures a translation's alignment with collective interpretations by computing the Euclidean distance between its affective vector and the affective centroid of all translations. Second, the Translator's Affective Profile (TAP), a visualization model that intuitively displays a translator's distinctive tendencies across affective dimensions, including positivity, negativity, neutrality, polarity, and subjectivity. We applied this framework to four English translations of the Chinese literary masterpiece Fusheng Liuji. Results effectively quantify significant disparities in affective delivery across translations and generate highly distinguishable translator affective profiles. This study offers new data-driven tools for translation quality assessment and demonstrates the potential of computational linguistics in uncovering translator subjectivity and stylistic signatures.