Diachronic Shifts in Dunhuang Translation Norms: A Corpus-Based Study of Cultural Diplomacy

This study presents a diachronic investigation of translation norms in English-language Dunhuang cultural heritage texts, drawing on the Dunhuang Cultural Heritage Parallel Corpus (DCHPC), which comprises 751,523 Chinese characters and 506,850 English words across 15,379 aligned units (14,121 body-text units) spanning 2010–2022. Grounded in Toury’s norm theory and Koskinen’s institutional translation framework, the study tracks shifts across two operational dimensions: cultural terminology strategies (D1) and the use of Sanskrit-restored forms (D3). A stratified sample of 400 occurrence-level items was coded by Coder A, with an independently coded stratified 100-item subset used to assess inter-coder reliability (Cohen’s κ = 0.921); 398 items entered the main analysis. Results show that pinyin transliteration (TR) was the dominant strategy in both periods (53.77% overall), though its proportion decreased from 58.0% in Period 1 (2010–2013) to 51.7% in Period 2 (2021–2022). Semantic translation (SEM) and Sanskrit restoration (SKT_RESTORED) both approximately doubled in the BRI Era (SEM: 6.1%→11.6%; SKT_RESTORED: 6.1%→10.1%). However, these descriptive trends did not reach statistical significance (χ ² (6, N = 398) = 8.97, p = .175, Cramér’s V = .150; permutation test, p = .172). Notably, the results do not support the widely assumed narrative of a rapid, dramatic normative shift driven by policy change; they instead provide a documented, reproducible empirical baseline for this domain. The study contributes a replicable coding methodology and offers evidence-based implications for heritage translator training.