聞分館君 Ask BunKanKun

唐書直筆

7 min read

Text ID
KR2o0005
Date
1044–1060
Dynasty
宋

Tángshū zhíbǐ 唐書直筆

Plain-Brush Notes for the New Tang History

by 呂夏卿 (Lǚ Xiàqīng, fl. 1042–1080s)

About the work

The Tángshū zhíbǐ is the surviving editorial workbook of one of the seven principal compilers of the KR2a0027 Xīn Tángshū 新唐書 — Lǚ Xiàqīng’s running argument, set down during the compilation period (1044–1060), about how the new history should be structured: which old-history passages were redundant or erroneous, what new categorical and rubric decisions were called for, and what the work’s promulgated fánlì 凡例 (compilation guidelines) ought to look like. The first three juàn discuss the revision of Annals (běnjì 本紀), Biographies (lièzhuàn 列傳), and Treatises (zhì 志), with sustained attention to the prolixities and lacunae of Liú Xù’s 劉昫 Jiù Tángshū; the fourth juàn is a Xīnlì xūzhī 新例須知 (“New conventions: things to be aware of”) — Lǚ’s draft of the compilation rubrics. According to Cháo Gōngwǔ’s 晁公武 Dúshū zhì 讀書志, this is what Lǚ argued for at the compilation bureau; some of his recommendations were adopted by Ōuyáng Xiū and Sòng Qí, others rejected.

Tiyao

We respectfully submit that Tángshū zhíbǐ was composed by Lǚ Xiàqīng of the Sòng. Lǚ, zì Jìnshū 縉叔, was a native of Jìnjiāng in Quánzhōu (Fújiàn). Promoted jìnshì, he was Sheriff of Jiāngníng, then served as Director of the Imperial Library Xuāndé láng shǒu mìshū chéng (宣德郎守秘書丞). On the completion of the Xīn Tángshū compilation, in which he had taken part, he was promoted to Vice-Director of the Imperial Library (zhí mìgé 直秘閣) and Senior Compiler of the Court of Imperial Sacrifices (tóng zhī lǐyuàn 同知禮院); later he went out to Prefect of Yǐngzhōu, where he died in office. His career is given in his Sòngshǐ biography. Examining Zēng Gōngliàng’s 曾公亮 memorial presenting the Xīn Tángshū, the seven listed compilers place Lǚ sixth. His biography says Lǚ was “long in historiography, deep-versed in Táng matters, having broadly culled traditions and zájì (informal accounts) from several hundred sources, balancing and ordering them; well-versed also in genealogy, originator of the genealogical Biǎo (tables); his contribution to the Xīn Tángshū was the most consequential.” So although his substantive rank was below those of Ōuyáng Xiū and Sòng Qí, his actual editorial labour was in no respect less.

According to Cháo Gōngwǔ’s Dúshū zhì, this book is what Lǚ proposed at the bureau: the first two juàn discussing Annals, Biographies, and Treatises; the third juàn listing the prolixities and lacunae of the Old History; the fourth juàn being the Xīnlì xūzhī — i.e., the proposed fánlì (compilation rubric). The Cháo recension records “Tángshū zhíbǐ in 4 juàn; Xīnlì xūzhī in 1 juàn”; the present recension consolidates them as 4 juàn total — perhaps a later combination. Cháo says Lǚ’s recommendations were partly adopted, partly not, by Ōuyáng and Sòng Qí; what is not perfectly aligned with the eventual Xīn Tángshū are precisely the points Ōuyáng and Sòng did not adopt. But “the man who praises red rejects white; the man who relishes sweet despises bitter” — every author goes his own way, and what is taken up is not necessarily right, and what is rejected not necessarily wrong.

We further note that Cháo Gōngwǔ separately records Lǚ’s Bīngzhì 兵志 in 3 juàn — said to have been obtained from Yǔwén Xūzhōng 宇文虛中. Jì Méng 季蒙 wrote at the end: “Lǚ in compiling the Táng history separately wrote a Bīngzhì in three sections, which he kept secret and ordered his children not to transmit. Bào Qīnzhǐ 鮑欽止 of the Personnel Bureau, fond of book-collecting, painstakingly sought it out; his son Wúwèi tàishǒu Gōngsūn happened to mention it, and we earnestly borrowed and copied it at the Mountain Studio in Wúxīng” — so Lǚ had distinct views on the Táng beyond the Bīngzhì of the Xīn Tángshū, but unlike Wú Zhěn’s 吳縝 open attack on the Xīn Tángshū (cf. Xīn Tángshū jiūmiù 新唐書糾謬), Lǚ’s were more guarded. The Bīngzhì is now lost; we preserve this work as comparative material for cross-collation. Qiánlóng 47, 5th month, respectfully revised by Jǐ Yún, Lù Xīxióng, Sūn Shìyì; reviser Lù Fèichí.

Abstract

Lǚ Xiàqīng was jìnshì in Qìnglì 2 (1042), and was recruited to the Xīn Tángshū compilation by Ōuyáng Xiū not long after the project was constituted in Qìnglì 4 (1044). His name appears sixth in Zēng Gōngliàng’s presentation memorial of Jiāyòu 5 (1060). His principal contributions to the published work were the great Zǎixiàng shìxì biǎo 宰相世系表 (j. 71–75 of the Xīn Tángshū) — the largest medieval Chinese genealogical compilation extant in any standard history — and substantial drafting of the Bīngzhì 兵志 (j. 50). The Sòngshǐ (j. 331) records his ranking expertise in genealogy (pǔxué 譜學) and his command of the corpus of informal Táng accounts.

The Tángshū zhíbǐ is therefore an exceptional document: a working compiler’s set of arguments preserved alongside the published history they helped produce. It was already partly out of date by the time of Jiāyòu 5: as the WYG tiyao notes, what does not perfectly map onto the Xīn Tángshū are the points where Ōuyáng and Sòng Qí declined Lǚ’s recommendations. The work is thus an essential comparator for Xīn Tángshū studies, complementing Wú Zhěn’s much later (and openly hostile) KR2a0028 Xīn Tángshū jiūmiù 新唐書糾謬. Lǚ kept his independent Bīngzhì 兵志 in 3 juàn secret in his lifetime, transmitting it only after considerable cajoling from Bào Qīnzhǐ — that work is now lost.

The juan-count discrepancy between Cháo Gōngwǔ’s “4 + 1” and the received “4” is best explained as the Xīnlì xūzhī having been folded into juàn 4 of the Zhíbǐ in late Sòng or early Yuán transmission; the contents-line in the WYG indeed reads “卷四：新例須知”. The work was little circulated under the late Sòng and Yuán; the Sìkù compilers reconstructed it from the Yǒnglè dàdiǎn and Bào Shìgōng’s 鮑士恭 family copy, with Cháo Gōngwǔ’s Dúshū zhì and the Wénxiàn tōngkǎo as cross-references. CBDB id 3649 records Lǚ as a Northern Sòng Qìnglì-era jìnshì but does not record his lifedates. Modern reference works (e.g. Quán Hànshēng 全漢昇) place his floruit roughly 1042–c. 1080, with death probably in the late 1070s or early 1080s.

Translations and research

No complete English translation located.

  • Charles Hartman, The Making of Song Dynasty History (Cambridge UP, 2021), §3.4 on the Xīn Tángshū compilers and their distinctive editorial signatures.
  • Denis Twitchett, The Writing of Official History under the T’ang (Cambridge UP, 1992), Ch. 7 on the Sòng reception of Táng historiography.
  • Liú Tōngshēng 劉統勝, “Lǚ Xiàqīng yǔ Xīn Tángshū de zǎixiàng shìxì biǎo” 呂夏卿與《新唐書》的宰相世系表, Zhōngguó shǐ yánjiū 中國史研究 (1998).
  • Yáo Yǒngzhì 姚永植, “Sòng Xīn Tángshū xiūzuàn kǎo” 宋《新唐書》修纂考, Wénxiàn 文獻 (2003).
  • Cài Chóngbǎng 蔡崇榜, Sòngdài xiūshǐ zhìdù yánjiū 宋代修史制度研究 (Wénjīn, 1991), Ch. 5.
  • Yú Yīngshí 余英時, Sòng Míng lǐxué yǔ zhèngzhì wénhuà 宋明理學與政治文化 (Yǔnchén, 2004), passim.

Other points of interest

The Tángshū zhíbǐ is one of the very few mid-Northern-Sòng historical-criticism texts to survive that is not primarily a moral catalogue (in the Tángjiàn / Lùnduàn mode) but rather an editorial-philological apparatus. Together with KR2a0028 Xīn Tángshū jiūmiù 新唐書糾謬 it forms the documentary backbone for any reconstruction of how the Xīn Tángshū was actually built — and unlike Wú Zhěn’s later attack, Lǚ’s notes are by an insider working in real time.

Graph View

Backlinks