A New Perspective on the Language of the Qur’an through AI-Assisted Linguistic Analysis

Abstract

This study examines how conceptual relationships in the language of the Qur’an can be analyzed through artificial intelligence–assisted distributional semantic methods. While meaning in classical Arabic linguistics and Qur’anic exegesis is interpreted through root-based morphology, contextual coherence (siyāq–sibāq), and syntactic harmony (naẓm), whether these principles can be represented through computational models constitutes a significant methodological question. The research employs digital Qur’anic corpora, primarily the Quranic Arabic Corpus, to conduct morphological analysis, semantic similarity measurement, and contextual modeling. Using a pre-trained AraBERT model, the semantic proximity between selected concepts (nūr, ẓulumat, and ʿadl) is calculated through cosine similarity metrics. The findings indicate that conceptual associations intuitively identified by classical exegetes can be quantitatively observed in distributional linguistic patterns. However, the results also demonstrate that AI-based analyses remain limited in capturing the theological and aesthetic dimensions of meaning. By bridging classical semantic theories with contemporary computational linguistics, this study proposes a methodological framework that contributes to the literature in Qur’anic studies and the digital humanities.