聞分館君 Ask BunKanKun

歸潛志

16 min read

Text ID
KR3l0079
Date
1235–1241
Dynasty
金-元

Guīqián zhì 歸潛志

Records of the Retreat-in-Hiding by 劉祁 (撰)

About the work

A fourteen-juàn historical bǐjì by 劉祁 Liú Qí 劉祁 (1203–1250; zì Jīngshū 京叔, of Húnyuán 渾源 in modern Shānxī), son of the Jīn censor Liú Cóngyì 劉從益. Guīqián — “retreat-in-hiding” — was the name Liú Qí inscribed over the door of his rebuilt study at Húnyuán when, in the rénchén year (1232) of the Mongol breach of Biànjīng 汴京 and the Jīn dynasty’s collapse, he fled north through Wèi, Qí, and Yān (some 2,000 lǐ) back to his ancestral village; the book takes its name from that study. Composed in the years following his return (the author’s zìxù is dated 1235; later layers extend through the early 1240s), it is by general consensus the single most important first-hand source for late-Jīn intellectual, literary, and political history, and for the Mongol conquest of Biàn in 1233–1234. Liú Qí witnessed the fall of Biànjīng personally — juàn 11 is his eyewitness narrative of the Āizōng’s flight, the surrender of the city, and the betrayal of Cuī Lì 崔立 — and the Jīnshǐ compilers under 脫脫 Tuōtuō drew on this work for the entire Āizōng běnjì. Juàn 1–6 are xiǎo zhuàn (brief biographies) of late-Jīn figures; juàn 7–10 mixed historical reports; juàn 11 the Lù Dàliáng shì (Record of Affairs at Dàliáng) on Āizōng’s fall; juàn 12 the Lù Cuī Lì bēi shì (on the coerced Cuī Lì stele inscription); juàn 13 a biànwáng (analysis of the dynasty’s fall) plus yǔlù-style miscellany; juàn 14 his own shīwén.

Tiyao

Your servants report: Guīqián zhì in 14 juàn, by the Yuán Liú Qí. Qí zì Jīngshū, of Húnyuán; son of the Censor [Liú] Cóngyì. As a Tàixué shēng (Imperial Academy student) he sat for the jìnshì examination but did not pass. When the Yuán troops entered Biàn, he fled back to his native village. In the wùxū year (1238) he came out again and was placed first in the Nánjīng selection-examination; he was appointed Shānxī Dōnglù kǎoshì guān (Examination Officer for the Eastern Circuit of Shānxī); he was then taken on by the staff of the Cóngzhēng Nánxíngshěng (Punitive Southern Branch Secretariat); after seven years he died. The old [bibliographies] placed him in the Wényì zhuàn of the Jīnshǐ and accordingly titled him a “Jīn man” — but this is not at all the truth. The book is called Guīqián because, in the rénchén (1232) northward return, Qí inscribed these two characters over his study, and so used them as the title of what he composed. Yet his later coming-out — his Western Hills’ integrity (Xīshān zhī jié) not maintained to the end — likewise is not the truth. The volume’s head has Qí’s own preface dated yǐwèi (1235), which says that what he had formerly heard and seen, in idle days he recorded as memory recovered them — taking down whatever came. From juàn 1 to juàn 6 are all brief biographies of late-Jīn persons; juàn 7 to juàn 10 are miscellaneous notes on surviving affairs (the Tíyào here writes “juàn 7 to juàn 30”, evidently a copyist’s slip for juàn 10 — the work is 14 juàn in total). Juàn 11 is titled Lù Dàliáng shì, recording the beginning and end of Āizōng’s loss of the dynasty. Juàn 12 is titled Lù Cuī Lì bēi shì, recording how when [Cuī] Lì rose in revolt the court officials erected a stele to flatter him, and coerced Qí to compose the inscription. Another piece is titled Biànwáng (Analysis of the Fall), narrating how the Jīn forebears had brought order, and how the late generation had brought disorder and ruin. From these two pieces down to juàn 13, all is miscellaneous notes, roughly in the form of yǔlù (recorded sayings), in no way matching what precedes. We suspect that these two pieces (Lù Cuī Lì bēi shì and Biànwáng) originally formed a single juàn at the tail of the work, with the yǔlù as juàn 13 and the shīwén as juàn 14 appended at the end; later transcribers, finding the piān uneven in length, cut the yǔlù in half and shifted part of it forward to fill out the present juàn 12, and so the structure stands askew.

In the rénchén upheaval Qí was at Biànjīng and witnessed events first-hand; what he records hits the truth. Therefore the Jīnshǐ in his own biography praises this zhì as “much sufficient to vouch for the affairs of the late Jīn”; the Āizōng běnjì takes its account wholly from what Qí wrote. Moreover: where the DàJīn guózhì says the Shūmìshǐ Yīlā Púā 伊喇蒲阿 went out to surrender to the Yuán, this zhì does not record the surrender — agreeing with the Jīnshǐ and proving the DàJīn guózhì’s error. Where the Yuánshǐ says that in the first month of rénchén (1232) Tàizōng crossed the [Yellow] River southward from Báibō and Ruìzōng forded the Hàn northward from Qiàoshítān — putting both crossings in the same time — this zhì records Ruìzōng’s forging of the Hàn as in the eleventh month of xīnmǎo (1231), and Tàizōng’s crossing of the [Yellow] River as in rénchén, agreeing with the Jīnshǐ and with the Mùān jí of Yáo Suì 姚燧 and the Míngchén shìlüè of Sū Tiānjué 蘇天爵 — proof of the Yuánshǐ’s error.

Again: it records that in Tiānxīng 1 (1232) Liú Yuánguī 劉元規 was sent as envoy to the Northern Court and his fate was unknown — the Jīnshǐ does not record this. It records that Sàkèsū 薩克蘇 acted as go-between to ruin the Yuánfēi (Imperial Consort) of the Lǐ clan — the běnjì does not give her name. It records that on the 1st of the 3rd month of Dàdìng 17 (1177), the envoys of various states having audience, rain occurred and the court was dismissed — agreeing with Zhōu Huī 周煇’s Běiyuán lù — but the běnjì records only the dismissal on the eclipse of the 1st of the 3rd month of Dàdìng 16. It records that the Jīn paper-currency law was changed in name eight times — the Jīnshǐ Shíhuòzhì fails to record that Tōnghuò was changed to Tōngbǎo, and Tōngbǎo changed back to Tōnghuò — all useful to supplement omissions in the official history. Then: where the Jīnshǐ Jiāopìnbiǎo says that in Dàdìng 16 (1176) the Sòng Tāng Bāngyàn 湯邦彥 served as Shēnqǐngshǐ (Asking-leave Envoy), this zhì writes Qíqǐngshǐ (Praying-request Envoy). The Túkètǎn Wūdēng zhuàn (= Púchá Wúdài 蒲察吾帶) says that in the first month of Tiānxīng 1 the court, hearing the [Yuán] troops had entered the Ráofēng Pass, transferred Wūdēng to head the Xíngshěng at Wénxiāng to guard Tóngguān — this zhì records the matter in Zhèngdà 8 (1231). The Wányán Sīliè zhuàn records that Wáng Wò 王渥 followed [Wányán] Sīliè to die in battle — this zhì writes that he followed Chíjiā Kāqíkā 持嘉喀齊喀. The [Wányán] Yīng zhuàn says that, encountering Yuán troops at Bàzhōu, [Yīng] was defeated and died — this zhì writes “encountered at Lù”. The Guō Ālín zhuàn says the Sòng troops arrived in mass and [Guō] thereupon fought and died — this zhì writes that his horse fell, he was taken captive, and it is unknown whether he lived or died. The Shī Ānshí zhuànzàn says he died in anger over a remonstrance — this zhì says that once in office his standing was much diminished — all giving discordant words. Many other discrepancies of date-order, surname, and rank with the official history. All can serve as material for mutual collation. Among those who treat of the Jīnyuán (Jīn sources’) surviving affairs, this zhì and 元好問 Yuán Hǎowèn’s Rénchén zábiān (壬辰雜編) are the most outstanding — the Jīnshǐ equally praises both. The Rénchén zábiān being lost, this zhì is all the more precious.

The current-circulating versions are all 8 juàn; even the Chuánshì Lóu stored copy is so. Our dynasty’s [Qīng] Guō Cháoyú’s 郭朝釪 Jīn shī, in compiling his anthology, only drew on the first 7 juàn — clearly he had not seen the complete text. This [present] copy in 14 juàn agrees with Wáng Yùn’s 王惲 Húnyuán shìdé bēi (Húnyuán generations’-virtue stele inscription) and may indeed derive from the Yuán-period transmitted copy. Qián Zēng’s 錢曾 Dúshū mǐnqiú jì states that Lù Mèngfú’s 陸孟鳬 family hand-copy of Guīqián zhì is in 14 juàn — that is presumably this copy. Respectfully checked, Qiánlóng 46 (1781), 8th month. Chief Compilers: Jì Yún 紀昀, Lù Xīxióng 陸錫熊, Sūn Shìyì 孫士毅. Chief Collator: Lù Fèichí 陸費墀.

Abstract

Liú Qí (CBDB 28855; 1203–1250) was born at Húnyuán 渾源 (modern Shānxī, Dàtóng prefecture) into the long-established Liú lineage that had served the Jīn through several generations. His grandfather Liú Wéi 劉撝 was a jìnshì and held office; his father 劉從益 Liú Cóngyì 劉從益 served as Imperial Censor under Xuānzōng and Āizōng. Liú Qí was raised in his father’s official residences south of the Yellow River, including Nánjīng (Kāifēng), where the late-Jīn court was based after the abandonment of Zhōngdū (Běijīng) in 1214. He studied with the leading Jīn literati of the Zhèngdà — Tiānxīng generation, including 趙秉文 Zhào Bǐngwén, 李純甫 Lǐ Chúnfǔ, and the slightly older 元好問 Yuán Hǎowèn — all three of whom appear as principal subjects in juàn 1–6 of the Guīqián zhì. He sat the Tiānxīng 1 (1232) jìnshì examination at Biànjīng but did not pass; that same year the Mongol siege of Biànjīng began.

The composition of the Guīqián zhì arose from Liú Qí’s experience of the 1232–1234 catastrophe. He was inside Biànjīng through the spring of 1233 — Āizōng’s flight on the 11th day of the 12th month of Tiānxīng 1 (early 1233), the regent Wányán Nǔshēn’s failed defence, the surrender to the Mongol general Sùbùtái 速不台 in the 4th month of Tiānxīng 2 (May 1233), and most notoriously the Cuī Lì coup, in which the Jīn general Cuī Lì 崔立 murdered the regents, surrendered the city to the Mongols, and coerced Liú Qí himself to draft the self-laudatory Cuī Lì bēi (a forced authorship Liú Qí would defend in juàn 12 of the present work as written under duress, the text deliberately ambiguous so that it could be repudiated). After the surrender Liú Qí was deported with the rest of the Biàn literati to the Mongol-occupied north; the zìxù of the work records that he travelled “from Wèi past Qí and into Yān, in all two thousand lǐ” before reaching Húnyuán in jiǎwǔ (1234), at age 32. There he built a study he named Guīqián — “the retreat-in-hiding” — and began the present work; the preface is dated the full-moon of the late summer of yǐwèi (1235).

The work is at once a memorial to the dead and an apologia for the survivor. Juàn 1–6 are biographical sketches (mostly two to ten lines each) of late-Jīn literati, officials, and military men — including extended treatments of Zhào Bǐngwén, Lǐ Chúnfǔ, Yuán Hǎowèn, Wáng Ruòxū 王若虛, Lǐ Jīng 李經, Liú Qí’s own father, and dozens of less prominent figures — most of whom had died in the JīnYuán upheaval. These biographies are the single most important source for the Zhèngdà (1224–1232) literary world of Biànjīng, the world of Liú Qí’s intellectual formation. Juàn 7–10 are mixed notes on Jīn court affairs, examination, ritual, and intellectual debate, including substantial passages on Jīn paper-currency reform, on the Dàoxué controversy at the Jīn court, and on the rise of Quánzhēn Daoism (the Wáng Zhé 王嚞 — Mǎ Yù 馬鈺 — Qiū Chǔjī 丘處機 lineage, important for Yuán-period religious history). Juàn 11, the Lù Dàliáng shì (Record of Affairs at Dàliáng), is Liú Qí’s eyewitness narrative of Āizōng’s flight from Biànjīng in late 1232, the surrender of the city, the betrayal of Cuī Lì, and the Mongol deportation of the Jīn literati — the single most important first-hand source for the fall of Biànjīng, and the principal source-base of Jīnshǐ Āizōng běnjì. Juàn 12, the Lù Cuī Lì bēi shì, is Liú Qí’s own defence of his coerced inscription. Juàn 13 contains the Biànwáng essay analysing the dynasty’s collapse (factional struggle at court; the alienation of the Jurchen aristocracy from the Sinicised bureaucracy; military overextension) plus a yǔlù section of philosophical and literary discussion. Juàn 14 contains his own shīwén.

The dating of the work has been examined by several modern scholars. The zìxù is dated 1235; but internal references mention events through the early 1240s (Liú Qí’s return to office in 1238 — when he placed first in the Nánjīng selection-examination under the new Mongol-controlled administration — and his subsequent service on the Cóngzhēng nánxíngshěng through to about 1245). The work was therefore composed in layers c. 1235–1241 and revised later; this is the bracket used for the catalog frontmatter here. The Sìkù’s irritation that Liú Qí’s “Western Hills’ integrity was not maintained to the end” (i.e., he eventually accepted Mongol office) is a MíngQīng moralism that does not affect the historical value of the work.

The text-history is complex. The work was widely transmitted in an 8-juàn truncation (which omits juàn 11–14, i.e., the most historically valuable portion — the eyewitness Biànjīng narrative, the Cuī Lì bēi defence, the Biànwáng, and the shīwén); all Míng-era printed editions and most Qīng-private-collection holdings (including the famous ChuánshìLóu of Chángshú) had only 8 juàn. The 14-juàn Sìkù version was reconstructed from a Yuán-period transmission preserved in Lù Mèngfú’s family hand-copy (cited via Qián Zēng’s Dúshū mǐnqiú jì) and from internal cross-collation with Wáng Yùn’s stele inscription for the Húnyuán Liú lineage. The 14-juàn form is the basis of all modern critical editions. Catalog meta classes the work under “元” (Yuán); but Liú Qí is a JīnYuán transitional figure, and the work — composed in the years immediately after the Jīn fall, treating overwhelmingly of Jīn matter, and considered the principal Jīn historiographic source after the lost Rénchén zábiān of Yuán Hǎowèn — is here classed in the frontmatter as JīnYuán transitional.

Standard modern edition: Cuī Wényìn 崔文印, coll., Guīqián zhì (Zhōnghuá shūjú, 1983, in the Yuán Míng shǐliào bǐjì cóngkān series). The Cuī Wényìn collation is the scholarly standard; it includes a detailed qiányán on text-history and is the basis of all subsequent translations and citations.

Translations and research

  • Chan, Hok-lam. 1970. The Historiography of the Chin Dynasty: Three Studies. Steiner (Wiesbaden). The foundational Western-language study; pp. 121–88 are devoted to Guī-qián zhì and its use by the Jīn-shǐ compilers. See Wilkinson §63.4.
  • Chan, Hok-lam. 1993. The Fall of the Jurchen Chin: Wang E’s Memoir on Ts’ai-chou under the Mongol Siege (1233–1234). Steiner. Repeated comparison with Liú Qí’s parallel account.
  • Chan, Hok-lam, and Wm. Theodore de Bary, eds. 1982. Yuan Thought: Chinese Thought and Religion under the Mongols. Columbia. Multiple chapters draw on Guī-qián zhì.
  • West, Stephen H. 1985. “Mongol Influence on the Development of Northern Drama.” In China under Mongol Rule, ed. John D. Langlois Jr. (Princeton). Uses Guī-qián zhì for the Mongol-period Yān-jīng literary scene.
  • West, Stephen H. 2001. “Crossing Over: Huizong in the Afterglow, or the Deaths of a Troubling Emperor.” In Emperor Huizong and Late Northern Song China, ed. Patricia Buckley Ebrey and Maggie Bickford (HUP). Uses Guī-qián zhì on the Jīn memory of Huī-zōng.
  • Tao, Jing-shen 陶晉生. 1976. The Jurchen in Twelfth-Century China: A Study of Sinicization. UWP. Foundational use of Guī-qián zhì for Jurchen acculturation.
  • Bol, Peter K. 1987. “Seeking Common Ground: Han Literati under Jurchen Rule.” HJAS 47.2: 461–538. Major analytical use of Guī-qián zhì for the Zhèng-dà literary world.
  • Hymes, Robert P. 2021. “A Tale of Two Sieges: Liu Qi, Li Gao, and Epidemics in the Jin-Yuan Transition.” Journal of Song-Yuan Studies 50: 293–365. The most recent major Western-language study; uses Guī-qián zhì to reconstruct disease history of the 1232–1234 siege.
  • Franke, Herbert. 1994. “The Chin Dynasty.” In The Cambridge History of China, vol. 6 (CUP). Repeated citation of Guī-qián zhì throughout the chapter.
  • Cuī Wén-yìn 崔文印. 1983. qián-yán to the Zhōnghuá edition. The standard modern Chinese-language introduction; provides exhaustive text-history.
  • No complete European-language translation of the Guī-qián zhì exists; substantial sections are translated in Chan (1970) and Hymes (2021).

Other points of interest

The Cuī Lì bēi (Cuī Lì stele) incident is one of the most-discussed cases of literary coercion in pre-modern Chinese history. Cuī Lì, having murdered the Jīn regents and surrendered Biànjīng to the Mongols in the 4th month of 1233, forced the captive Hànlín literati to compose a stele extolling his “merit”. Wáng Ruòxū 王若虛 refused outright; the task fell to Liú Qí, who composed an inscription ambiguous enough to be later disavowed. Juàn 12 of the Guīqián zhì is Liú Qí’s own defence of the inscription — pleading the impossibility of refusing without execution, and arguing that the surviving text contains coded indictments of Cuī Lì that any educated reader would recognise. Modern scholarship (Hok-lam Chan; Stephen West) generally accepts Liú Qí’s defence; the inscription itself does not survive (Cuī Lì was killed by the Mongols a few months after the surrender and the stele was almost certainly destroyed).

The Biànwáng essay in juàn 13 is one of the earliest and most analytically sophisticated examples of the wángguó (dynasty-collapse) reflective essay genre, antedating Wáng Fūzhī’s Sòng lùn by four centuries and influencing the later genre directly. Its principal argument — that the Jīn fell not for lack of military strength but because of the alienation of the Jurchen aristocracy from the Hàn-Chinese bureaucracy and the consequent breakdown of the central state’s administrative coherence — is now standard in Jīn historiography.

Graph View

Backlinks