RECONCEPTUALIZING LITERARY GENRE THROUGH STYLOMETRIC ANALYSIS: A COMPUTATIONAL LINGUISTIC APPROACH TO TEXT CLASSIFICATION

DR. MAHWISH SHAMIM, MARIAM AKRAM, DR. IRAM RUBAB, AYESHA INAM, DR. MUHAMMAD ARFAN LODHI

Background: This study rethinks literary genre from a linguistic and computational perspective, challenging the traditional view of genre as a fixed and rule-bound classification system. Drawing on structuralist and post-structuralist insights, genre is conceptualized as both systematic and fluid. Within this context, stylometry is positioned as a linguistic methodology that quantifies stylistic variation through lexical, syntactic, semantic, structural, and phonological features, enabling a shift from interpretive to empirical analysis.

Method: The study adopts a theoretical–computational approach rather than empirical text analysis. It develops a stylometric framework for modeling genre across poetry, drama, and fiction by transforming texts into multidimensional feature vectors. Comparative feature taxonomy is constructed, and computational techniques including statistical methods, machine learning, and deep learning are critically evaluated. A hybrid modeling strategy is proposed to integrate interpretability with predictive capability.

Discussion: The findings demonstrate that genre distinctions emerge from configurations of linguistic features rather than isolated markers. Poetry is characterized by phonological and structural constraints, drama by dialogic and interactional patterns, and fiction by narrative and syntactic complexity. Based on these insights, the study proposes the Multidimensional Stylometric Genre Model (MSGM), in which genres are represented as probabilistic clusters within a multidimensional linguistic feature space. The MSGM further incorporates a hybrid analytical mechanism that combines interpretable statistical features with machine learning-based pattern recognition, addressing the tension between explainability and accuracy.

Conclusion: The study concludes that genre should be understood as a dynamic, probabilistic linguistic construct rather than a rigid category. The MSGM offers a flexible, scalable, and theoretically grounded computational model that aligns with contemporary digital humanities practices and supports more nuanced genre analysis across literary forms.