Shū jīng jízhuàn 書經集傳

Collected Commentary on the Classic of Documents by 蔡沈 (zhuàn 撰)

About the work

The single most influential Sòng commentary on the Shàngshū 尚書 (KR1b0001), composed at Zhū Xī’s 朱熹 express deathbed commission and serving as the standard Shàngshū gloss of the Yuán–Míng–Qīng examination system. Cài Shěn 蔡沈 (Jiǔfēng 九峯, 1167–1230), Zhū Xī’s most trusted late disciple, was charged with the project in winter Qìngyuán 5 / 1199, the winter before Zhū Xī died (1200); he completed the manuscript in spring Jiādìng 2 / 1209 — ten years to the project, almost exactly. Zhū Xī had already established the redaction of the “two Diǎn” 二典 (Yáo diǎn, Shùn diǎn) and the “three ” 三謨 (Dà Yǔ mó, Gāo Yáo mó, Yì Jì) before his death; Cài Shěn arbitrated all the rest. The work canonically circulates in 6 juǎn (the four-dynasty division: 1 juǎn for Yú, 1 for Xià, 1 for Shāng, 3 for Zhōu — a hundred chapters in original conception, of which 58 survive); a 1-juǎn annotated Xiǎo xù 小序 (in which Cài Shěn dismantles the traditional chapter-prefaces, paralleling Zhū Xī’s own treatment of the Máo prefaces in the Shī jízhuàn) and a 1-juǎn Zhū Xī wèndá 朱熹問答 (preserving Cài Shěn’s correspondence with Zhū Xī on disputed readings) accompanied the original presentation copy of 12 fascicles submitted to court in the Chúnyòu era (1241–1252) by Cài Shěn’s son Cài Háng 蔡杭. The Wèndá is now lost; the Xiǎo xù survives though commercial reprints excise it. The text became the imperially prescribed Shàngshū commentary in the Yuán examination reform of 1313, was reaffirmed in the Míng Wǔjīng dàquán 五經大全 (1415), and remained so until the Qīng — the Sìkù tíyào (submitted Qiánlóng 41 / 1776) deliberately frames it as the standard against which all rival Sòng Shū commentaries are measured.

Tiyao

Imperially Authorized Sìkù Quánshū. Classics, division 2. Shū jízhuàn. Books-class.

Précis. Your servants etc. respectfully submit: the Shū jízhuàn in six juǎn is by Cài Shěn of the Sòng. Shěn, zì Zhòngmò, hào Jiǔfēng, was a man of Jiànyáng — son of [Cài] Yuándìng. His personal record is appended to the Sòngshǐ biography of [Cài] Yuándìng. In Qìngyuán jǐwèi (1199) Master Zhū commissioned Shěn to compose the Shū zhuàn, and by Jiādìng jǐsì (1209) the book was complete. — (Note: this is according to the year-and-month of the autograph preface. Zhēn Déxiù, in writing Shěn’s epitaph, says “after some decades it was at last brought to completion” — this is clearly an error caused by an extraneous “decades” character.) — During the Chúnyòu era his son [Cài] Háng presented the work to court, declaring it as a Jízhuàn in six juǎn, Xiǎo xù in one juǎn, Zhū Xī wèndá in one juǎn, transcribed into twelve fascicles. The Wèndá fascicle has long been lost; Dǒng Dǐng’s Shū zhuàn zuǎnzhù 書傳纂注 records that “the Chúnyòu palace-presentation copy contained Master Zhū’s letters to Cài Zhòngmò and several segments of Yǔlùn”; these are now categorically inserted into the Gānglǐng jí lù 綱領輯錄 of [Dǒng Dǐng’s book], so the Wèndá’s text still survives scattered through Dǒng’s work, even though the original section-arrangement can no longer be reconstructed.

The Xiǎo xù in one juǎn — Shěn likewise refutes the [traditional chapter-prefaces] line by line, just as Master Zhū attacked the Shī xù — its text still survives, but the bookseller’s editions all excise and do not print it. According to Zhū Shēng’s Shàngshū pángzhù: “The old-text Shū xù originally formed a single chapter; the Kǒng commentary moved it to head each chapter individually; Mr Cài cut it out and placed it at the back, in order to preserve the older form.” This was, presumably, the principle Master Zhū had passed down. — (Note: Chén Zhènsūn’s Shūlù jiětí records: “Master Zhū’s [edition of the] ancient classic, four juǎn; preface, one juǎn.” If so, this is the redaction Master Zhū had himself fixed — meaning that there was already a complete book, and Zhū Shēng’s claim that this was Master Zhū’s transmitted principle is presumably an oversight.) — Yuán-end and Míng-early printings still kept the Xiǎo xù attached. Yet what the Sòngshǐ yìwénzhì records is also only six juǎn, so it would seem that from the Sòng down [the Xiǎo xù] was already published as a separate piece, while the Jízhuàn circulated alone. Hé Yìsūn 何異孫 of the Yuán in his Shíyī jīng wèn duì 十一經問對 says: “The Cài zhuàn cut at Jízhōu still has the Shū xù prefixed to each chapter; this in no way harms the Cài zhuàn” — i.e. it is one workshop’s plate and not a general convention.

Shěn’s preface says that the “two Diǎn” 二典 and “three ” 三謨 had been diǎndìng 點定 by Master Zhū. But Dǒng Dǐng, on the entry “zhèng yuè shuò dàn” (first month, first day), notes: “Master Zhū’s own compilation of the Shū zhuàn extends only as far as the Kǒng preface; for the rest of the great points he transmitted [his readings] orally to Mr Cài, together with autograph drafts of more than a hundred passages, which Mr Cài was to use to bring the work to completion” — meaning that the Dà Yǔ mó itself was not yet fully finished by Master Zhū; the preface’s “two Diǎn, three ” is only an approximate summary. Dǒng further quotes Chén Lì 陳櫟: — (Note: this entry of Chén’s is not contained in his Shū zhuàn zuǎnshū; it must be from his Shū zhuàn zhézhōng.) — “Master Zhū’s revised draft once read: ‘zhèng yuè — i.e. of the next year. As for shénzōng 神宗, [the various] commentators take it as referring to Shùn’s enthroned ancestors Zhuānxù and Yáo, hence treating shénzōng as the temple of Yáo — uncertain whether this is right.‘” Phrases like “dì zhī chū” (the original [enthroning] of the Lord) etc. — Master Zhū had not affirmatively pronounced these to be the temple-of-Yáo readings, whereas the present text says so unambiguously: was this Master Zhū’s own later revision, or was it altered by Mr Cài? Then what the preface calls “diǎndìng by Master Zhū” likewise cannot have escaped some emendation. Hence at the close of Sòng, Huáng Jǐngchāng 黃景昌 and others each wrote Zhèng wù biàn yí 正悞辨疑 (“Correcting Errors and Discriminating Doubts”) on this work; and Chén Lì, Dǒng Dǐng, Jīn Lǚxiáng 金履祥 — all of whom faithfully studied Master Zhū’s learning — wrote, respectively, the Shū zhuàn zhézhōng, the Shū zhuàn zuǎnshū, and the Shàngshū biǎozhù, all of them rigorously argumentative on the points where the Cài zhuàn requires correction. Under the Hóngwǔ-era Xiūshū zhuàn huì xuǎn 修書傳會選 of the Míng, sixty-six items were corrected.

— Our reigning Court — its imperially established Shū jīng zhuàn shuō huìzuǎn 書經傳說彙纂 has likewise made many corrections and refinements [to the Cài]: in Master Zhū’s view, Shàngshū exegesis should “comprehend what is comprehensible and leave alone what is not comprehensible” (tōng qí suǒ kě tōng ér quē qí suǒ bù kě tōng 通其所可通而闕其所不可通) — a position appearing repeatedly in his Yǔlèi; whereas Shěn, on the Yīn pán and Zhōu gào, demands an explanation for every passage one by one. That this should leave him not without regret is only natural. Yet his exposition is comparatively simple and easy, and his sources are sound — the larger character is in the end pure. Under the Yuán it stood as the standard alongside the Old Commentaries (see the Yuánshǐ Xuǎnjǔ zhì), and people set the Old Commentaries aside and studied this book; under the Míng it stood as standard alongside Xià Shàn’s xiángjiě (see Yáng Shèn’s Dānqiān lù), and again people set Xià’s xiángjiě aside and studied this book — there is good reason for that. Respectfully submitted, Qiánlóng 41 / 1776, tenth month.

— Director-General, Jì Yún 紀昀, Lù Xīxióng 陸錫熊, Sūn Shìyì 孫士毅. — Director of Final Collation, Lù Fèichí 陸費墀.

Abstract

The Shū jīng jízhuàn is the dominant Shàngshū commentary of late-imperial China. It was commissioned by Zhū Xī 朱熹 of his closest Shū-disciple, Cài Shěn 蔡沈 (the second son of Cài Yuándìng 蔡元定, 1135–1198, Zhū Xī’s principal Yìjīng-collaborator), in the winter of Qìngyuán 5 / 1199 — that is, on the cusp of Zhū Xī’s death. The chronology is that given in Cài Shěn’s own autograph preface, dated Jiādìng jǐsì sānyuè jì wàng — Jiādìng 2 / 1209.III.16, at Wǔyí 武夷. The frontmatter window 1199–1209 covers the entire decade.

This dating contradicts an extraneous claim in Zhēn Déxiù’s 真德秀 epitaph for Cài Shěn that the work took “shù shí nián” 數十年 (“some decades”); the Sìkù compilers explicitly correct this as a textual slip in the epitaph (a stray shí 十 (“ten”) character) — a rare case where the tíyào explicitly arbitrates a catalog-vs-external dating discrepancy in favor of the work’s own preface against the xíngzhuàng / 墓誌. The actual division of labor is: the “two Diǎn” (Yáo, Shùn) and the Dà Yǔ mó of the “three ” had been provisionally vetted by Zhū Xī himself before his death (though, per Dǒng Dǐng’s 董鼎 Shū zhuàn zuǎnzhù, the Dà Yǔ mó was not in fact fully finished); the rest is Cài Shěn’s own composition, executed on the basis of Zhū Xī’s oral transmission and over a hundred manuscript drafts in Zhū Xī’s hand.

The Chúnyòu-era (1241–1252) palace-presentation by Cài Shěn’s son Cài Háng 蔡杭 includes three components: (1) the Jízhuàn proper, in 6 juǎn, organized by the four-dynasty division (1 juǎn Yú, 1 juǎn Xià, 1 juǎn Shāng, 3 juǎn Zhōu) covering the 58 surviving chapters of the post-Qín-fire Shàngshū; (2) the annotated Xiǎo xù 小序 in 1 juǎn, in which Cài Shěn — following Zhū Xī’s treatment of the Máo shī xù in the Shī jízhuàn — refutes the traditional chapter-prefaces line by line; (3) the Zhū Xī wèndá 朱熹問答 in 1 juǎn, preserving Zhū Xī’s letters and Yǔlèi fragments specifically on disputed Shàngshū readings. The Wèndá fascicle is irrecoverably lost (preserved fragmentarily in Dǒng Dǐng’s Zuǎnzhù); the Xiǎo xù survives but is excised from commercial reprints. The 6-juǎn recension circulating from the Sòng Yìwén zhì through to the present is the Jízhuàn alone.

The work has a long sub-tradition of “correction” literature precisely because of the precarious authority of those passages where Cài Shěn extended Zhū Xī’s unfinished readings: Huáng Jǐngchāng’s 黃景昌 Zhèng wù biàn yí, Chén Lì’s 陳櫟 Shū zhuàn zhézhōng 書傳折衷, Dǒng Dǐng’s 董鼎 Shū zhuàn zuǎnshū 書傳纂疏, and Jīn Lǚxiáng’s 金履祥 Shàngshū biǎozhù 尚書表注 — all by orthodox ZhūXī partisans — each correct specific Cài readings, and the Hóngwǔ Míng Shū zhuàn huìxuǎn 書傳會選 institutionally corrected sixty-six items. The Qiánlóng-imperial Shū jīng zhuàn shuō huìzuǎn 書經傳說彙纂 (Kāngxī era) is the climax of this tradition.

The Sìkù compilers’ substantive critique is precise: Zhū Xī’s own Shū-method, repeatedly stated in the Yǔlèi, was “tōng qí suǒ kě tōng ér quē qí suǒ bù kě tōng” (comprehend what is comprehensible, leave a lacuna where it is not), but Cài Shěn departs from this by demanding a gloss for every passage of the famously opaque Yīn pán 殷盤 and Zhōu gào 周誥. Despite this departure the Sìkù treats the work as canonical, judging its prose lucid, its sources sound, and its larger character “in the end pure” — the closing line of the tíyào (under Yuán the Cài zhuàn set the Old Commentaries aside; under Míng it set Xià Shàn’s xiángjiě aside, KR1b0011) is in effect a survey of why the Cài commentary won and others lost.

Translations and research

A complete English translation of Cài Shěn’s Shū jízhuàn has not been published; James Legge’s The Chinese Classics III: The Shoo King (Hong Kong / London, 1865) translates the Shàngshū canon and frequently cites the Cài zhuàn but does not render it whole. Standard scholarship: Liú Qǐyú 劉起釪, Shàngshū yánjiū yàolùn 尚書研究要論 (Jǐnán: Qílǔ shūshè, 2007); Cài Gēnxiáng 蔡根祥, Sòngdài Shàngshū xué àn 宋代尚書學案 (Taipei: Huámùlán, 2006); Jīn Shēngyáng 金生楊, Cài Shěn 〈Shū jízhuàn〉 yánjiū 蔡沉《書集傳》研究 (Chéngdū: Bā-Shǔ shūshè, 2008) — the last is the standard monograph. For the Yuán-Míng-Qīng curricular reception see Benjamin A. Elman, A Cultural History of Civil Examinations in Late Imperial China (Berkeley: University of California Press, 2000), and Hìlde de Weerdt, Competition over Content: Negotiating Standards for the Civil Service Examinations in Imperial China (Cambridge, MA: Harvard Asia Center, 2007). For Zhū Xī’s Shàngshū method specifically, see Tsē-tsung Chow [Chow Tse-tsung], Wén-lín 文林 (Madison: University of Wisconsin Press, 1968), and the relevant chapter in Daniel K. Gardner, Chu Hsi: Learning to Be a Sage (Berkeley, 1990).

Other points of interest

The most consequential single feature of the work for the textual history of the Shàngshū is Cài Shěn’s open acceptance of the Gǔwén Shàngshū 古文尚書 chapters as authentic — including the Dà Yǔ mó — together with his elevation of the Dà Yǔ mó phrase rén xīn wéi wēi, dào xīn wéi wéi, wéi jīng wéi yī, yǔn zhí jué zhōng 人心惟危、道心惟微、惟精惟一、允執厥中 (“The human mind is precarious, the dào mind is subtle; refine it, unify it, hold genuinely the mean”) to the status of a xīnfǎ 心法 — a sixteen-character “transmission of the mind” — of the sage-kings, traced through Yáo–Shùn–Yǔ. This sixteen-character formula, almost certainly a Hàn-period interpolation, was thereby installed at the heart of Lǐxué 理學 self-understanding for the next six hundred years; Yán Ruòqú’s 閻若璩 Shàngshū gǔwén shūzhèng 尚書古文疏證 (1745) and the entire mid-Qīng kǎojù 考據 demolition of the Gǔwén Shàngshū are in large measure an attack on the credibility-base of this work.

The Hóng fàn 洪範 commentary in the Jízhuàn is conspicuously thin; Cài Shěn’s separate Hóng fàn huáng jí nèi piān 洪範皇極內篇 (5 juǎn, also in Sìkù) was where he developed the Hóng fàn number-cosmology project that his father Cài Yuándìng had pioneered.

The Jízhuàn is also one of the best-attested cases of Sòng commentarial succession: we know exactly when, by whom, and to what unfinished state Zhū Xī handed off the project, and we have both the Jízhuàn preface and Dǒng Dǐng’s auxiliary documentation of the handover.