An Interpretable Stylometric Framework for Early Chinese Buddhist Translator Styles: A Case Study on the Attribution of Disputed Works of An Shigao
Organizations: Department of Information Management, Peking University · Research Center for Digital Humanities, Peking University · Department of Philosophy and Religious Studies, Peking University
Abstract
Attributing early Chinese Buddhist translations requires evidence that scholars can trace back to the texts. We develop an interpretable stylometric framework that uses features defined or reviewed by Buddhist philologists, including function characters, sentence formulas, sentence-final characters, sentence length, and part-of-speech n-grams. We construct stylistic profiles from 31 authenticated translations by seven translators from the Eastern Han through the Western Jin, then compare eight disputed texts with An Shigao's profile. Two texts, T1557 and T0101, show the strongest stylistic similarity to the authenticated An Shigao corpus, with scores of 0.783 and 0.687, respectively. The attributions agree with those reported by a published BERT-based approach on the same set of disputed texts. The framework requires no task-specific training and allows scholars to inspect which linguistic features support each proposed attribution.
Figures & tables
| Translator | Title (English) | ID | Volumes | Sentences | Characters |
|---|---|---|---|---|---|
| Yan Fotiao | Six Pāramitās Sutra | T0778 | 1 | 52 | 1069 |
| An Shigao | Long Āgama on Ten Dharmas | T0013 | 2 | 574 | 10090 |
| Sūtra on the Desire for Human Life | T0014 | 1 | 529 | 6255 | |
| Sūtra on the Four Noble Truths | T0031 | 1 | 101 | 1858 | |
| Sūtra on the Roots of Suffering | T0032 | 1 | 239 | 3655 | |
| Sūtra on the Four Truths of Suffering | T0036 | 1 | 32 | 729 |
| Title (English) | CBETA ID | Volumes | Sentences | Characters |
|---|---|---|---|---|
| Saṃyuktāgama Sutra | T0101 | 1 | 428 | 9247 |
| Sūtra on the Five Aggregates | T0105 | 1 | 51 | 787 |
| Dhammacakkappavattana Sutta | T0109 | 1 | 32 | 770 |
| Thirty-Seven Factors of Enlightenment | T0604 | 1 | 16 | 931 |
| Sūtra on Meditation and Contemplation | T0605 | 1 | 10 | 281 |
| Sūtra on Receiving the Dharma | T0792 | 1 | 17 | 304 |
| Feature Type | Characters (Chinese) | Transliteration |
|---|---|---|
| Sentence-final Char | 也, 是, 之, 乎, 者, 故, 何, 行, 提, 法 | ye, shi, zhi, hu, zhe, gu, he, xing, ti, fa |
| Template | Example (Chinese) | Function (English) |
|---|---|---|
| 第 X | 第一、第二… | Sequential enumeration |
| X 者 | 色者、心者… | Definition or exemplification |
| 以 X 为 Y | 以心为界、以苦为道 | Categorical relation |
| 何等故举 | 何等、何者、云何 | Interrogative / doctrinal question |
| 佛言 | 佛言:… | Quotation formula |
| 已 X 便 | 已灭便得道 | Temporal / conditional sequence |
| Category | Function Words (Chinese) | English Gloss |
|---|---|---|
| Pronouns | 之, 其, 自, 皆, 是 | pronouns (of, its, self, all, to be) |
| Prepositions | 以, 於, 为, 与, 及 | prepositions (by, in, for, with, and) |
| Particles | 所, 者, 故 | particles (that which, one who, therefore) |
| Conjunctions / Adverbs | 而, 有, 则, 无, 不, 如, 能 | conjunctions/adverbs (and, have, then, none, not, like, can) |
| Feature | Description | p_holm | ||
|---|---|---|---|---|
| func_之_per100 | Frequency of particle zhi (之,nominalizer) per 100 chars | 0.838 | ||
| func_其_per100 | Frequency of pronoun qi (其,“his/her/its”) per 100 chars | 0.798 | ||
| pos3_v_u_n_prop | POS trigram Verb–Auxiliary–Noun proportion | 0.749 | ||
| func_皆_per100 | Frequency of adverb jie (皆, “all”) per 100 chars | 0.738 | ||
| func_及_per100 | Frequency of conjunction ji (及, “and ”) per 100 chars | 0.735 | ||
| pos3_d_v_d_prop | POS trigram Adverb–Verb–Adverb proportion | 0.724 | ||
| Author | Scripture | func_之 _per100 | func_其 _per100 | func_皆 _per100 | func_及 _per100 | pos2_ v_d _prop | pos2_ n_r _prop | pos3_ v_u_n _prop | pos3_ d_v_d _prop | pos3_ n_u_v _prop | pos4_ n_v_r_n _prop |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Yan Fotiao | T0778 | 41.10% | 28.47% | 46.27% | 29.34% | 37.14% | 31.10% | 50.81% | 39.18% | 24.77% | 23.18% |
| An Shigao | T0013 | 29.07% | 28.47% | 31.49% | 31.06% | 71.76% | 28.34% | 30.59% | 74.50% | 23.47% | 27.04% |
| T0014 | 29.07% | 28.47% | 31.54% | 29.34% | 56.83% | 37.28% | 27.95% | 63.44% | 30.22% | 34.96% | |
| T0031 | 29.07% | 28.47% | 30.85% | 29.34% | 89.42% | 47.01% | 22.98% | 94.63% | 29.36% | 56.40% | |
| T0032 | 29.07% | 29.46% | 32.40% | 29.34% | 82.59% | 23.21% | 26.95% | 72.50% | 27.32% | 26.67% | |
| T0036 | 29.07% | 28.47% | 30.85% | 29.34% | 66.03% | 20.82% | 22.98% | 60.89% | 17.11% | 23.18% |
| Author | func_之 _per100 | func_其 _per100 | func_皆 _per100 | func_及 _per100 | pos2_ v_d _prop | pos2_ n_r _prop | pos3_ v_u_n _prop | pos3_ d_v_d _prop | pos3_ n_u_v _prop | pos4_ n_v_r_n _prop |
|---|---|---|---|---|---|---|---|---|---|---|
| Yan Fotiao | 41.10% | 28.47% | 46.27% | 29.34% | 37.14% | 31.10% | 50.81% | 39.18% | 24.77% | 23.18% |
| An Shigao | 29.19% | 28.64% | 32.80% | 30.42% | 70.70% | 31.71% | 28.43% | 71.88% | 29.58% | 31.73% |
| Kang Senghui | 87.55% | 68.88% | 42.92% | 41.92% | 31.83% | 53.82% | 86.78% | 28.18% | 58.39% | 73.03% |
| Kang Mengxiang | 41.99% | 50.75% | 51.49% | 85.06% | 27.17% | 61.31% | 44.61% | 22.94% | 79.05% | 84.40% |
| Lokakṣema | 44.70% | 75.10% | 67.07% | 60.41% | 37.19% | 76.01% | 63.22% | 42.91% | 70.80% | 56.79% |
| Zhi Qian | 66.36% | 59.04% | 65.91% | 49.57% | 32.82% | 53.02% | 63.96% | 36.53% | 64.53% | 60.49% |
| CBETA ID | Our Results | BERT model Results |
|---|---|---|
| T0101 | Yes | Yes |
| T0105 | No | No |
| T0109 | No | No |
| T0604 | No | No |
| T0605 | No | No |
| T0792 | No | No |
Authorship Attribution
Authorship attribution identifies or verifies who wrote a text, code sample, or other creative work from patterns in its style and content.
See all 71 papers