Borrowing the Canonical Form: Formal Anchoring and Selective Semantic Drift in Late Ming Commentaries on the 'Awakening of Faith'

Overview This replication package contains the dataset and computational pipeline for the paper "Borrowing the Canonical Form: Formal Anchoring and Selective Semantic Drift in Late Ming Commentaries on the ‘Awakening of Faith’" by Junzhe Yu. The project applies Word2Vec embeddings and Procrustes orthogonal alignment to quantify semantic drift and textual reuse within the Chinese Buddhist canon, specifically focusing on the commentary tradition of the Awakening of Faith (大乘起信論) during the Late Ming dynasty. Repository Structure To ensure strict data provenance (Human-in-the-Loop tracking), the data assets are managed in a tiered architecture within the provided archive: data/: Contains the text corpus structured in four tiers: 01_raw_archive: Unmodified ZIP archives downloaded from CBETA. 02_manual_interim: Texts with metadata and prefaces manually removed. 03_auto_clean: Pure Chinese character streams generated by Module 1. 04_manual_slices: Manually segmented subsets for document-level drift testing. code/: Contains Main_Analysis_Pipeline.ipynb, an interactive computational workflow. outputs/: Contains auto-generated semantic drift tables, textual reuse rates, Leave-One-Out (LOO) robustness checks, and pre-trained Word2Vec models. Reproducibility To reproduce the findings presented in the paper: Ensure your environment matches the snapshot in the included requirements.txt. Execute the Jupyter Notebook (code/Main_Analysis_Pipeline.ipynb) in a strictly linear sequence. Note: A strict random seed (seed=42) and single-thread execution (workers=1) are enforced within the code to ensure the exact replication of the vector space and distance metrics. For more details, please refer to the Reproducibility_Statement.md file included in the archive. License and Citation The code and datasets are released under the MIT License (see the LICENSE.txt file). If you utilize this dataset or pipeline in your research, please cite it using the metadata provided in the CITATION.cff file or via the Zenodo citation tool on this page.