聞分館君 Ask BunKanKun

史通

12 min read

Text ID
KR2o0001
Date
710–722
Dynasty
唐

Shǐtōng 史通

Comprehensive Treatise on Historiography

by 劉知幾 (Liú Zhījǐ, 661–721)

About the work

The Shǐtōng is the earliest extant systematic Chinese treatise on historiography — a work of historical criticism (shǐpíng 史評) by a working official historian, written outside his bureau duties. It is divided into ten juàn of “Inner Chapters” (Nèipiān 內篇, originally 39 sections of which 36 survive) discussing the principles, forms, and craft of writing history, and ten juàn of “Outer Chapters” (Wàipiān 外篇, 13 sections) treating the history of historical writing itself and offering critical evaluations of past historians. Liú himself dated the inner block to Jǐnglóng 4 (710); the outer chapters were completed posthumously by his second son 劉餗 (Liú Sù) on the basis of Liú’s drafts, finished by 722. The title is modelled on Bān Gù’s 班固 Báihǔ tōng 白虎通, with a side-glance at the lost Hàn-era proposal to ennoble a descendant of Sīmǎ Qiān as Shǐtōng zǐ 史通子. The work is the first to attempt a systematic typology of historical genres (the famous “Six Schools” 六家 and “Two Forms” 二體), the first to demand explicit canons of evidence and source criticism, and the first to challenge the privileged status of the Confucian classics as historical witnesses — for which it was widely vilified down to the Qing.

Tiyao

The Shǐtōng, twenty juàn, was composed by Liú Zǐyuán of the Táng. Zǐyuán’s birth-name was Zhījǐ; to avoid the tabooed name of Míng huáng (Emperor Xuánzōng) he went by his courtesy name. He was a native of Péngchéng. At his capping he was promoted jìnshì; he was assigned as Sheriff of Huòjiā, then advanced to Drafter of the Phoenix Pavilion (Fènggé shèrén 鳳閣舍人), concurrently editing the dynastic history. Under Zhōngzōng he was promoted to Crown Prince’s Coachman (Tàizǐ shuàigèng lìng 太子率更令), and after a series of advancements to Imperial Library Director (Mìshū jiān 祕書監), Crown Prince’s Senior Attendant (Tàizǐ zuǒ shùzǐ 太子左庶子), and Academician of the Chóngwén Hall. At the beginning of the Kāiyuán era he reached the rank of Standing Attendant (Zuǒ sǎnqí chángshì 左散騎常侍); later, implicated in a case, he was demoted to Deputy Prefect of Ānzhōu, and died in office. His career is given in his biography in the Tángshū.

This book was completed in the fourth year of Jǐnglóng [710]. There are ten Inner juàn in thirty-nine sections, and ten Outer juàn in thirteen sections. He wrote it during his tenure as Imperial Library Director, having quarrelled with Xiāo Zhìzhōng 蕭至忠, Zōng Chǔkè 宗楚客, and others over historical matters and not seeing eye to eye — therefore he composed the book out of indignation. Three of the Inner-chapter sections — Tǐtǒng 體統, Pìmiù 紕繆, and Chízhāng 弛張 — survive only as titles; consulting the original biography in the Tángshū, which already says that the Shǐtōng contains forty-nine sections, the loss of these three predates the Sòng compilation of the Xīn Tángshū.

The Inner Chapters discuss historiographical form and the determination of right and wrong; the Outer Chapters narrate the lineage of historical works and offer miscellaneous evaluations of past men’s strengths and shortcomings. The text occasionally repeats material between the two divisions, occasionally contradicts itself. Opening the Inner Liù jiā 六家 chapter, it begins by saying “the writings of the rulers of antiquity have been treated in detail in the Outer Chapters” — meaning the Outer Chapters were written first, and the Inner Chapters were a distillation, with the original draft material not entirely deleted.

Zǐyuán was deeply learned in historiography, served in the History Office for nearly thirty years, and was at the bureau longer than anyone. His command of past and present, and his diagnosis of historical writing’s strengths and weaknesses, are something later men cannot match. But his temperament was excessively sharp, his diction often inflamed, and his denunciations went too far — sometimes he would not be reined in. The “Doubting Antiquity” (Yígǔ 疑古) and “Confounding the Classics” (Huòjīng 惑經) chapters have been universally condemned, and need not detain us. Even in the Liù jiā chapter, he criticises the Shàngshū for irregular precedent; in Zàiyán 載言 he criticises Zuǒ shì for not following ancient method; in Rénwù 人物 he criticises the Shàngshū for omitting the Eight Worthies (Bā Yuán 八元) and Eight Talents (Bā Kǎi 八愷), Hán Zhuó, Fēi Lián, È Lái, Hóng Yāo, and Sàn Yí Shēng, and the Chūnqiū for omitting Yóu Yú, Bǎilǐ Xī, Fàn Lǐ, Wén Zhǒng, Cáo Mò, Gōngyí Xiū, Níng Qī, and Ráng Jū — also a particular absurdity.

As to the historian’s craft, what is at stake is praise and blame, not titles. Even the most benighted and tyrannical rulers — the Yōu and the Lì — cannot have their kingly title struck off. Yet his Chēngwèi 稱謂 chapter would have all the temple-titles of the Jìn rulers from Kāng and Mù onwards effaced. Zhū Yún 朱雲’s snapping of the railing and Zhāng Gāng 張綱’s burying of the wheel are paragons of unwavering rectitude — yet his Yányǔ 言語 chapter dismisses them as petty matters that the historian should not bother to record. Qú Yuán 蘧瑗 (Qú Bóyù) was rewarded with a senior post and never withdrew into seclusion, yet his Pǐnzǎo 品藻 chapter complains that the Gāoshì zhuàn 高士傳 omits him. Confucius’s disciples wished to honour Yǒuruò 有若 — this is recorded in the Mèngzǐ, and is certainly no fabrication — yet his Jiànshí 鑒識 chapter, where the Shǐjì records the same incident, calls Sīmǎ Qiān cruder than Chǔ Shàosūn. All of this is a wilful raising up and casting down — extraordinarily one-sided.

In the Záshuō 雜說 chapter he points to Zhào Dùn’s “fish meal” as not actually frugal fare — gainsaying the Gōngyáng — and says “bamboo horses” are not native to Bīngzhōu, attacking the Hòu Hàn jì; these are likewise petty and meandering. Moreover, the Zhōulǐ gives the Grand Astrologer charge of the Six Codes of state and the Lesser Astrologer charge of the records of the various states — meaning that the historian was responsible for institutional and cultural matters as well; this is the ancient pattern. But Zǐyuán took praise-and-blame as the sole purpose of historiography, and treated everything else as mere appendage; he therefore rejected the Biǎo 表 and Zhì 志 of the Shǐjì and Hànshū with their treatises and tables, mostly demanding their excision — most contrary to ancient practice.

He chides the Hòu Hànshū for accepting miscellaneous accounts, but himself relies on the Zhúshū jìnián 竹書紀年 and the Shānhǎi jīng 山海經. He chides the Hànshū “Treatise on Five Elements” for errors, but himself takes Yuán Huī’s 元暉 Kē lù 科錄 to have been written by Yuán Huīyè 暉業 of the Wèi Jǐyīn princes; takes the biography of Liú Yú in the Hòu Hànshū to be in the Sānguó zhì — small inattentions which he could not avoid. Yet his anatomising and dividing — laying out a thing as black against white, so that with one extraction even Sīmǎ Qiān or Bān Gù would have nothing to say in self-defence — may indeed be called the legalist of historical method, the auditor of historical writing.

Since the Ming, three or four annotated editions have appeared, but their corruptions and dislocations are uniform throughout. The present base text is the old block-edition held in the imperial household (Nèifǔ), without notes, and seems closer to the ancient state than the others. The “Diǎn fán” 點煩 chapter has lost its red punctuation in every recension, and so it remains here; nothing can be supplied by collation, and so we leave it.

Abstract

Liú Zhījǐ (the catalog form 知幾 reflects the Sòng restoration; in his own lifetime he wrote and signed himself by his courtesy name 子玄, which was further altered to 子元 under the Qīng to avoid the Kāngxī taboo on 玄) wrote the Shǐtōng during his long career inside the Táng History Office (Shǐguǎn 史館), where he served on and off for some thirty years from 702 onwards. According to his autobiographical Zìxù 自序 (Nèipiān 36) and his celebrated 708 letter of resignation to Xiāo Zhìzhōng 蕭至忠 (preserved as Wàipiān 13), the project grew out of frustration with collective compilation, with the political pressure on historians by sitting officials, and with the obstruction of source access — the five complaints he there elaborates. The Inner Chapters (the theoretical core) were finished in 710. After Liú’s death in 721, his second son 劉餗 (Liú Sù) used his father’s drafts to complete the Outer Chapters by 722.

The work survives in its near-complete form (three of the original 39 inner sections, Tǐtǒng, Pìmiù, and Chízhāng, were already lost by the time of the Xīn Tángshū’s compilation in the eleventh century, as the Sìkù tiyao notes). The first known printing is the 1535 Lù Shēn 陸深 edition; before that the work circulated in manuscript for the better part of eight centuries. The standard annotated edition is KR2o0002 Shǐtōng tōngshì 史通通釋 by 浦起龍 (Pǔ Qǐlóng, 1679–c. 1762), printed in 1750, which incorporates and extends the earlier annotations of Lǐ Wéizhēn 李維楨, Guō Kǒngyán 郭孔延, Wáng Wéijiǎn 王維儉, and others. Modern punctuated editions usually rest on Pǔ; the convenient 2008 Shànghǎi gǔjí press edition prints Pǔ’s notes alongside Lǚ Sīmiǎn’s 呂思勉 1934 commentary as side-marginalia.

The Shǐtōng is famously hostile to canonical authority. Its Yígǔ 疑古 (“Doubting Antiquity”) and Huòjīng 惑經 (“Confounding the Classics”) chapters identify factual contradictions between the Shàngshū, Lúnyǔ, Mèngzǐ, Zuǒzhuàn, and Shǐjì and use them to argue that the classical accounts of Yáo, Shùn, Tāng, and Wén cannot all be true; they were widely censured from the Sòng onward, and the Sìkù editors above represent the conservative reception. Equally controversial were Liú’s structural complaints — that biǎo 表 (tables) and zhì 志 (treatises) should be cut from dynastic histories, that single-dynasty composite annals-biography (jìzhuàn 紀傳) form is preferable to general history (tōngshǐ 通史) — and his demand that the historian be a zhí bǐ 直筆 (straight-brushed) recorder rather than a praise-and-blame moralist.

The book lays out two influential schemata: the Six Schools (Liù jiā 六家) of historical writing — the Shàngshū, Chūnqiū, Zuǒshì, Guóyǔ, Shǐjì, and Hànshū — and the Two Forms (Èr tǐ 二體) — biānnián 編年 (chronological / annalistic) and jìzhuàn 紀傳 (composite annals-biography). These remained the standard analytical frame in Chinese historiography until the early twentieth century. Wilkinson (Chinese History: A New Manual, §46.6.4.1) treats the Shǐtōng as the foundational work of the East-Asian historiographical tradition and the principal point of comparison for KR2o0025 Wénshǐ tōngyì 文史通義 by 章學誠 (Zhāng Xuéchéng, 1738–1801), the Qing thinker who reopened many of Liú’s questions a thousand years later.

The local Kanripo source is from Sìbù cóngkān 四部叢刊 (SBCK), reproducing the 1602 Zhāng Dǐngsī 張鼎思 collation of the 1535 Lù Shēn print, with Jiāng Diànyáng’s 姜殿揚 supplementary collation notes (Shǐtōng zhájì bǔ 史通札記補).

Translations and research

A complete English scholarly translation now exists:

  • Victor Cunrui Xiong, tr., Shitong: A Thorough Exploration in Historiography, by Liu Zhiji. University of Washington Press, 2023. The first complete Western-language scholarly translation of both Inner and Outer Chapters.
  • Damien Chaussende, tr., Traité de l’historien parfait (Les Belles Lettres, 2014; selection). Annotated French translation of selected chapters.

Earlier partial translations:

  • William Hung, “A T’ang Historiographer’s Letter of Resignation”, Harvard Journal of Asiatic Studies 29 (1969): 5–52 — translation and annotation of the resignation letter (Wàipiān 13).

Substantial scholarship:

  • Nishiwaki Tsuneki 西脇常記, Shitsū no kenkyū 史通の研究. Kyoto: Dōhōsha, 1989. With companion translation volume, 2002. The standard Japanese scholarly edition.
  • Pǔ Qǐlóng 浦起龍, Shǐtōng tōngshì 史通通釋 (= KR2o0002). The standard premodern commentary, 1750.
  • Lǚ Sīmiǎn 呂思勉, Shǐtōng pínglín 史通評林 / commentary, 1934.
  • Hú Wéiyì 胡惟一 et al., Shǐtōng quányì 史通全譯, Guìzhōu rénmín, 1997. Modern Chinese translation.
  • Zhào Lǚfù 趙呂甫, Shǐtōng xīn jiàozhù 史通新校注, Chóngqìng chūbǎn shè, 1990. Modern critical edition with full apparatus.
  • E. G. Pulleyblank, “Chinese Historical Criticism: Liu Chih-chi and Ssu-ma Kuang”, in W. G. Beasley and E. G. Pulleyblank (eds.), Historians of China and Japan, Oxford UP, 1961, 135–166.
  • Tsuneo Masui, “Liu Chih-chi and the Shih-t’ung”, Memoirs of the Research Department of the Tōyō Bunko 34 (1976): 113–162.
  • Damien Chaussende, “Epilogue: Treatises according to Tang historian Liu Zhiji”, in Daniel Patrick Morgan and Damien Chaussende (eds.), Monographs in Tang Official Historiography, Brill, 2019, 343–357.
  • Xiaojing Miao, “‘Defying the Times’: Liu Zhiji’s Resignation Letter Reconsidered”, Tang Studies 40 (2022): 55–82.
  • On Hàn translation: On-cho Ng and Q. Edward Wang, Mirroring the Past: The Writing and Use of History in Imperial China, University of Hawai’i Press, 2005, ch. 5.

Other points of interest

The Shǐtōng’s influence is paradoxical: written by a permanent insider of the official-history apparatus, the work nevertheless survives largely outside the Twenty-Four Histories tradition it helped to systematise. Its proposal to abolish the biǎo 表 was generally ignored by later compilers; its taxonomies of the Six Schools and Two Forms were absorbed wholesale into Sìkù-period bibliographic vocabulary; its skepticism toward the classics was suppressed in mainstream Confucian transmission and rediscovered only when Qing evidential scholars began their own reinterrogation of the Wǔ jīng 五經.

Liú’s letter of resignation (the so-called “Five Complaints” of Wàipiān 13) is one of the most cited bureaucratic documents in pre-modern Chinese intellectual history; it is read both as a manifesto for the autonomy of the historian and as a working description of what was wrong with collective imperial historiography in the early Táng. Sòng-era readers from Liú Shù 劉恕 onwards generally agreed with the diagnosis even when they refused the larger anti-classical agenda.

Graph View

Backlinks