Computational Discovery of Chiasmus in Ancient Religious Text

TL;DR

This work addresses the challenge of detecting chiastic structures in Biblical texts by introducing the first computational framework that uses multilingual embeddings to capture lexical and semantic symmetry across half-verses and verses. It defines a formal chiastic score, $Final Score = \mu_{\text{chiasmus}} - \mu_{\text{non-pair}}$, and a significance criterion via $z_i > 3$, applied over sliding windows of length $N \in [4,8]$ with $z_i$ standardized against the distribution of scores. The approach is evaluated with top-50 candidates annotated by experts, achieving high inter-annotator agreement ($\kappa$ values of 0.76 and 0.89) and precision@k of 0.60 for half-verses and 0.80 for verses, identifying thousands of chiastic structures across Genesis, Numbers, and other books. The study demonstrates a scalable, interpretable method for chromatic literary analysis in ancient manuscripts and lays groundwork for a labeled chiastic corpus and broader cross-text applications.

Abstract

Chiasmus, a debated literary device in Biblical texts, has captivated mystics while sparking ongoing scholarly discussion. In this paper, we introduce the first computational approach to systematically detect chiasmus within Biblical passages. Our method leverages neural embeddings to capture lexical and semantic patterns associated with chiasmus, applied at multiple levels of textual granularity (half-verses, verses). We also involve expert annotators to review a subset of the detected patterns. Despite its computational efficiency, our method achieves robust results, with high inter-annotator agreement and system precision@k of 0.80 at the verse level and 0.60 at the half-verse level. We further provide a qualitative analysis of the distribution of detected chiasmi, along with selected examples that highlight the effectiveness of our approach.

Figures (1)

• Figure 1: Total number of chiasmi per Book at verse and half-verse level. While some books tend to have more chiasmi overall, this figure shows whether verse-level or half-verse-level chiasmi are more prevalent in each book.