尚書古文疏證

Shàngshū gǔwén shū zhèng 尚書古文疏證

Sub-commentary and Verification of the Old-Text Documents by 閻若璩 (zhuàn 撰)

About the work

The single most important Chinese philological monograph of the early-Qīng period: Yán Ruòqú’s 閻若璩 (1636–1704) demonstration that the 25 Gǔwén Shàngshū 古文尚書 chapters and the Kǒng Ān’guó zhuàn — accepted as authentic canonical text from the Eastern Jìn through YuánMíng — are in fact an Eastern-Jìn forgery. The work, in 8 juǎn, marshals 128 separate evidentiary arguments (tiáo 條) — citational, textual, institutional, geographic, chronological, philological — to establish the case. Its authority was decisive: by the early eighteenth century the Gǔwén forgery thesis was the philological consensus, despite Máo Qílíng’s 毛奇齡 vigorous counter-attack in his Gǔwén Shàngshū yuān cí 古文尚書冤詞 (“Lament-and-Plea for the Gǔwén Shàngshū”), and the Qing kǎojù tradition’s subsequent half-century of work treated the Shàngshū gǔwén shū zhèng as the founding methodological template.

The work had a complex composition history. Manuscript work began in the 1660s/1670s; the first four juǎn were complete and circulated under Huáng Zōngxī’s 黃宗羲 preface in Yán Ruòqú’s lifetime; the second four juǎn were drafted later and not all fully integrated; the work was not printed in full until well after Yán’s death. The transmitted text shows traces of unfinished editorial state: the third juǎn is partly lost; six entries (juǎn 2 nos. 28–30; juǎn 7 nos. 102, 108–110; juǎn 8 nos. 122–127) carry titles in the table of contents but lack full bodies; and the work as a whole is not perfectly arranged. The Sìkù tíyào explicitly identifies the surviving recension as “cǎo chuàng zhī běn” 草創之本 (a draft-form text).

Tiyao

Imperially Authorized Sìkù Quánshū. [Classics, division 2.] Gǔwén Shàngshū shū zhèng. [Books-class.]

Précis. Your servants etc. respectfully submit: the Gǔwén Shàngshū shū zhèng in eight juǎn — by Yán Ruòqú of our State. Ruòqú, zì Bǎishī, was a man of Tàiyuán, who relocated to Shānyáng. In Kāngxī jǐwèi (1679) he was nominated for Bóxué hóngcí [special examination].

The Gǔwén Shàngshū, compared to the jīnwén, has sixteen additional chapters; from JìnWèi onward there has been no school-based explanation [of these added chapters]. Therefore in the Zuǒ zhuàn citations Dù Yù annotates each as “lost-from-the-Shū.” At the start of the Eastern Jìn, the [Méi Zé recension] book first emerged, with twenty-five additional chapters; at first these still stood alongside the jīnwén. From Lù Démíng using it as the basis of his Shìwén, and Kǒng Yǐngdá using it as the basis of his Zhèngyì, [the gǔwén and the jīnwén] were merged with Fú Shēng’s twenty-nine into one [canon].

Although from the Tang onward there were canon-doubting and antiquity-doubting voices like Liú Zhījī’s, even Liú listed the Shàngshū as a “single school” in his Shǐ tōng and made no claim of gǔwén falsity. The first dissenting voices come from Wú Yù; Master Zhū too begins to entertain doubts; Wú Chéng and others, taking Zhū’s lead, sequentially extracted the forgery’s traces — the case became progressively clearer. Yet even they were not yet able to cut into the seams item by item to expose the gaps. Only with Méi Zhuó of the Míng (KR1b0038) does the first comparison-of-various-books-to-prove-its-pilfering occur; but his range of reading was somewhat narrow, his collection not exhaustive. Then with Ruòqú: he draws on canonical and ancient sources, one by one laying out each contradiction’s reason, and the gǔwén forgery is at last fully clarified. The 128 entries he sets out — Máo Qílíng wrote a Gǔwén Shàngshū yuān cí contesting them in a hundred ways, but in the end could not subdue the proper-principle by force-of-argument. Evidence-grounded words stand from the start undefeatable.

The book first took shape as four juǎn, with Huáng Zōngxī of Yúyáo writing a preface for it; the [further] four juǎn were drafted in continuation. After Ruòqú’s death, in transcription the third juǎn was lost; the second juǎn’s entries 28–30, the seventh juǎn’s entries 102 and 108–110, and the eighth juǎn’s entries 122–127 all have titles in the catalog but lack full bodies. The arrangement of the entries’ order has not yet been brought into discipline. Probably this is still a draft-form text.

In its body, places where occasionally not perfectly verified: e.g. relying on the Zhèngyì’s record of Zhèng [Xuán]‘s preface to the Shū, holding that what Mǎ-and-Zhèng transmitted does not match the chapter inventory of the Kǒng zhuàn — that argument is most exact. But the further claim that the Mǎ and Zhèng commentary recensions perished in the Yǒngjiā 永嘉 disorder is in fact incorrect. On investigation, the recensions of the two schools are still recorded in the Suí zhì, listed as commenting on twenty-nine chapters in total; the Jīngdiǎn shìwén fully cites them, also confined to twenty-nine — that is, removing the sixteen “no-school-explanation” chapters, leaving twenty-nine [Fú Shēng-equivalent], which matches Fú Shēng’s number. So [the MǎZhèng recension] is not in fact a separate-recension Kǒng-text commentary. Ruòqú mistakes Zhèng’s “lost”-list as the lost chapters Zhèng commented on — a single oversight in the midst of a thousand careful considerations.

Again: the Shǐjì and Hànshū both record only that [Kǒng] Ān’guó submitted the Gǔwén Shàngshū; there is no record of his receiving an imperial command to compose a commentary. This is a manifest proof of the falsity of [the Eastern-Jìn] received forged recension, and a key hinge for distinguishing the forgery — yet [Yán Ruòqú] sets it aside and does not say it. Also somewhat lacking-in-detail.

After several other entries he frequently extends into adjacent matters at length, often filling whole juǎn. Probably he was anxious that his Qiánqiū zhájì, separately authored, might not be transmitted; he therefore had it appended into [this work]. In the end this is digressive-and-tangled. Further: where the front juǎn argues a position, the back juǎn often refutes it, and he is unwilling to delete the front statement — although he follows the principle of Zhèng Xuán’s Lǐ jì commentary (first using Lǔ shī readings and not later revising), in editorial format this is still un-adjusted. Nevertheless, his back-and-forth detailing in order to lift the great millennial doubts — in the kǎo zhèng style of scholarship, this work has from the start no precedent. Respectfully submitted, Qiánlóng 43 / 1778, sixth 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 Shàngshū gǔwén shū zhèng is the single most consequential philological monograph in the history of Chinese canonical scholarship: Yán Ruòqú’s 閻若璩 (1636–1704) cumulative demonstration, in 128 separate evidentiary arguments organized into 8 juǎn, that the 25 chapters of the Gǔwén Shàngshū 古文尚書 and the accompanying Kǒng Ān’guó zhuàn are an Eastern-Jìn forgery (with the Méi Zé 梅賾 transmission of c. 320 CE marking the Gǔwén’s first appearance). The work decisively settled the Gǔwén authenticity question: from the early eighteenth century onward, the Gǔwén was treated by mainstream Chinese philology as forged, despite vigorous defenses by Máo Qílíng (in the Gǔwén Shàngshū yuān cí), Chén Dì (KR1b0043), and others. Modern scholarship continues to accept the basic forgery thesis, although later work has refined the chronology and authorship.

The composition window in the frontmatter (1670–1704) covers Yán Ruòqú’s mature career; the work was begun in the 1660s/1670s, completed in manuscript by the 1690s (with Huáng Zōngxī’s preface to the first four juǎn), and not fully printed until well after Yán’s death (the standard 1745 print). The Sìkù submission was Qiánlóng 43 / 1778.

The Sìkù tíyào on the work is exceptionally substantive — registering the kǎojù lineage from Wú Yù to Zhū Xī to Wú Chéng to Méi Zhuó to Yán Ruòqú; identifying Yán as the consummating figure (“kǎo zhèng zhī xué zé gù wèi zhī huò xiān yǐ” 考證之學則固未之或先矣 — “in the study of evidence-based verification, this work has from the start no precedent”); and giving an unusually detailed catalog of the work’s structural and substantive limitations.

The tíyào’s critical observations are precise and worth registering. (1) Structural: the work survives in draft form, with entries missing or incomplete in juǎn 2, 3, 7, and 8 — Yán Ruòqú had not finished editorial integration when he died, and the posthumous transmission preserves the unfinished state. (2) The Yǒngjiā argument: Yán Ruòqú claimed that the Mǎ Róng and Zhèng Xuán Shàngshū commentary recensions perished in the Yǒngjiā disorder of 311; the Sìkù compilers correct this — both recensions are still listed in the Suí zhì, and Lù Démíng’s Jīngdiǎn shìwén cites them on 29 chapters (= the 29 surviving chapters of the jīnwén). The MǎZhèng commentary therefore covered the jīnwén canon, not the additional gǔwén chapters Zhèng knew only as “lost.” (3) The “Kǒng Ān’guó received-imperial-command-to-compose-a-commentary” issue: neither the Shǐjì nor the Hànshū records this — i.e., the very institutional claim of the Kǒng zhuàn’s origin is prima facie false on the face of the Hàn historical record. The Sìkù compilers note that this is “the key hinge” of the forgery question, and that Yán Ruòqú surprisingly does not deploy it. (4) Structural-editorial: Yán Ruòqú often appends digressive material from his parallel Qiánqiū zhájì into the Shū zhèng — the compilers identify this as concern that the Zhájì might not be separately transmitted. (5) Front / back juǎn contradictions: where Yán changes his mind in later juǎn he does not edit out the earlier statement.

Despite these specific criticisms, the Sìkù verdict is decisive in Yán Ruòqú’s favor: the Shàngshū gǔwén shū zhèng is the founding monument of Qing kǎojù, and “no precedent” exists for its method.

The work’s relationship to the four immediately preceding Shū lèi entries — Méi Zhuó’s Kǎo yì (KR1b0038), Chén Dì’s Shū yǎn (KR1b0043), Huáng Dàozhōu’s Hóng fàn míng yì (KR1b0044), Yè Fāngǎi / Kùlènà’s Rì jiǎng (KR1b0045), and Wáng Fūzhī’s Bài shū (KR1b0047) — is implicit but pointed: it builds directly on Méi Zhuó’s case, refutes Chén Dì’s defense, supersedes Huáng Dàozhōu’s Hóng fàn number-cosmology by retiring the entire Gǔwén Shàngshū basis on which such Hé tú / Luò shū readings depended, and replaces the Kāngxī Rì jiǎng’s Cài-Shěn-orthodox doctrinal reading with a philological-evidentiary method. In the Sìkù Shū lèi sequence, the Shū zhèng is therefore the climactic transformation of the entire Càizhuàn-based commentary tradition.

Translations and research

The Shàngshū gǔwén shū zhèng is treated extensively in Western scholarship: Bernhard Karlgren, “The Authenticity of Ancient Chinese Texts,” BMFEA 1 (1929), and The Book of Documents (Stockholm: Östasiatiska Museet, 1950); Edward L. Shaughnessy, “Shàng shū” in Michael Loewe, ed., Early Chinese Texts: A Bibliographical Guide (Berkeley: Society for the Study of Early China, 1993); Benjamin A. Elman, From Philosophy to Philology: Intellectual and Social Aspects of Change in Late Imperial China (Cambridge, MA: Harvard Council on East Asian Studies, 1984), which treats Yán Ruòqú as the founding figure of Qing kǎojù. For the work’s specific arguments see Jiǎng Shànguó 蔣善國, Shàngshū zōngshù 尚書綜述 (Shanghai: Shànghǎi gǔjí, 1988); Liú Qǐyú 劉起釪, Shàngshū yánjiū yàolùn 尚書研究要論 (Jǐnán: Qílǔ shūshè, 2007). For Máo Qílíng’s defense of the Gǔwén see his Gǔwén Shàngshū yuān cí (separately in the Sìkù, KR1b0049).

Other points of interest

The Sìkù compilers’ explicit identification of the Kǒng Ān’guó received-command-to-compose-a-commentary claim as “the key hinge” of the forgery question — and their observation that Yán Ruòqú underplays it — is itself a methodological contribution: the compilers are doing additional kǎojù in the body of their own tíyào. This kind of constructive-additional argumentation is rare in the Sìkù tíyào sequence and indicates the editorial team’s serious engagement with Yán Ruòqú as their immediate scholarly predecessor.

The 128 entries — preserved in surviving form despite the unfinished state — set the genre template for Qing kǎojù: the cumulative-evidence monograph organized around discrete numbered arguments. Subsequent works in the same tradition (Wáng Mǐngshèng’s Shàngshū hòu àn, Sūn Xīngyǎn’s Shàngshū jīngǔwén zhù shū, Chéng Tíngzuò’s Shàngshū yǐ shù, etc.) all draw on this structural template. Yán Ruòqú’s Sì shū shì dì and his Qiánqiū zhájì are similarly structured.

Links

• CBDB: see 閻若璩 person note
• Wikidata: https://www.wikidata.org/wiki/Q11129418 (閻若璩)
• Sìkù quánshū zǒngmù tíyào, Shū lèi, Gǔwén Shàngshū shū zhèng entry (Kyoto Zinbun digital edition)