聞分館君 Ask BunKanKun

唐語林

8 min read

Text ID
KR3l0055
Date
1102–1110
Dynasty
北宋

Táng yǔlín 唐語林

Forest of Tang Sayings by 王讜 (撰)

About the work

An 8-juàn topical compendium of Táng anecdotes by 王讜 Wáng Dǎng 王讜 (zì Zhèngfǔ 正甫, of Chángān; c. 1050 – c. 1110), modelled directly on Liú Yìqìng’s 劉義慶 Shìshuō xīnyǔ 世說新語 (KR3l0001) but treating Táng-dynasty material. Wáng excerpted over fifty Táng xiǎoshuō — many of them now lost — and redistributed the material under the thirty-five Shìshuō headings (déxíng 德行, yányǔ 言語, zhèngshì 政事, wénxué 文學, etc.), to which he added seventeen new headings of his own (including shìhào 嗜好, “preferences”), making fifty-two mén in total. The work was originally in ten juàn but had largely disappeared by the Míng; the present 8-juàn form is the Sìkù compilers’ reconstruction from the surviving Míng-Jiājìng fragment of Qí Zhīluán 齊之鸞 and the Yǒnglè dàdiǎn 永樂大典 extracts.

Tiyao

Your servants report: Táng yǔlín in 8 juàn, by the Sòng Wáng Dǎng. Chén Zhènsūn’s Shūlù jiětí says: “Wáng Dǎng Zhèngfǔ of Chángān took fifty Táng xiǎoshuō, imitated the Shìshuō in dividing the material under thirty-five headings, and added a further seventeen headings, making fifty-two mén.” Cháo Gōngwǔ’s Jùnzhāi dúshū zhì says: “Compiler unknown; imitates the Shìshuō style, divided by mén, recording Táng-era míngyán (famous sayings), newly adding shìhào (preferences) and other seventeen headings, the rest as before.” Mǎ Duānlín’s Jīngjí kǎo cites Chén’s words under xiǎoshuō-school and Cháo’s words under zájiā-school — the two headings cross-list it, but it is in fact a single book. Only Chén’s version has 8 juàn and Cháo’s has 10 juàn — the numbers do not match; yet Chén further notes that the Guǎngé shūmù records 11 juàn lacking the jìshì and following 15 headings, while another edition has only 8 juàn with the headings all intact — these are differences arising from copyists’ divisions and combinations, hence the two recensions differ.

Dǎng’s name does not appear in the standard histories. Examining the entry on Péi Jí 裴佶 in the work, the character Jí is left blank with the note “imperial taboo-name”; this can only be the taboo of Sòng Huīzōng Zhào Jí, so Wáng Dǎng was a man of the Chóngníng — Dàguān era. Although this work imitates the Shìshuō, what it records is diǎnzhāng gùshí (institutional precedents) and jiāyán yìxíng (admirable sayings and excellent deeds), much of which illuminates the standard histories — unlike Liú Yìqìng, who specialised only in qīngtán (pure conversation). Moreover the various books he excerpted are mostly no longer extant, so his service in póují (collecting) is particularly not to be dismissed. Sadly, the printed edition has long been lost; thus the Míng Xiè Zhàozhè’s Wǔ zázǔ 五雜組 quotes Yáng Shèn’s 楊慎 remark that “Yǔlín is rarely transmitted and people scarcely know of it.”

Only the Wǔyīngdiàn shūkù 武英殿書庫 holdings have a defective edition cut in early Míng Jiājìng by Qí Zhīluán of Tóngchéng, divided into two juàn upper and lower, going from déxíng down to xiányuán — eighteen mén only — with Qí’s own preface stating that what he obtained was not a good copy: the characters and strokes were blurred and the chapter-order disordered, almost unreadable. We have now used what is preserved in the Yǒnglè dàdiǎn to collate cross-wise, deleting duplicates and adding over four hundred entries; and have obtained one piece of the original preface-and-table, recording the titles of books excerpted and the ménlèi general-table, so that the original editorial principle can still be recovered in outline — clearly at the start of the Míng the complete book still existed. But the Yǒnglè dàdiǎn entries are scattered under each successive rhyme, so the original ménmù (headings) are hard to recover by conjecture; we have provisionally ordered by shídài (period) and appended to the printed edition, with those of no fixed period following thereafter, making four juàn in total. And since the upper-and-lower 2-juàn of the printed edition have too many leaves per juàn, we have each juàn split into two, again making 8 juàn to restore the old form. This book has long had no collated edition, and étuō (errors and omissions) are many, the wényì (sense) often hard to follow; we have taken the Xīn and Jiù Tángshū and the various shuōbù (anecdote works) and carefully corrected — and where it is genuinely unknowable, have left the original reading, so as not to lose the principle of quēyí (leaving doubt). Respectfully collated and submitted, the tenth month of Qiánlóng 49 [1784]. Chief compilers: Jì Yún 紀昀, Lù Xīxióng 陸錫熊, Sūn Shìyì 孫士毅. Chief collator: Lù Fèichí 陸費墀.

Abstract

The Táng yǔlín is the principal Sòng-era Shìshuō-style topical anthology of Táng anecdotes, and a particularly valuable one because Wáng Dǎng (CBDB id 1889) excerpted over fifty Táng xiǎoshuō of which the majority are no longer extant — the work has the status of a near-Táng compendium of Táng zájì / yìshì literature, second only to Tàipíng guǎngjì 太平廣記 (KR3j0036) in importance for Táng anecdote material. Its dating to the Chóngníng – Dàguān period (1102–1110) is fixed by the internal Huīzōng taboo on Jí 佶 (the personal name of Zhào Jí, r. 1100–1126), as the Sìkù compilers observed. Wilkinson (Chinese History: A New Manual §62.3.11 #5) gives composition as c. 1106, with Wáng’s lifedates as c. 1050–1110.

Transmission was vexed. The Sòngshǐ Yìwén zhì recorded the work in ten juàn; later catalogues (Chén Zhènsūn, Cháo Gōngwǔ, Guǎngé shūmù) give variants of 8, 10, and 11 juàn, reflecting copyists’ divisions of the same 52-mén material. By the late Míng the book had become rare — Xiè Zhàozhè quotes Yáng Shèn complaining that “Yǔlín is rarely transmitted and people scarcely know of it.” The only surviving printed witness was a defective two-juàn edition cut by Qí Zhīluán of Tóngchéng 桐城 in early Jiājìng, covering only eighteen mén (déxíng through xiányuán); Qí himself acknowledged his exemplar was poor. The Sìkù compilers (1784) then collated Qí’s text against the scattered Yǒnglè dàdiǎn extracts, recovered over four hundred additional entries, retrieved one fragment of the original preface-and-mùlù, and rearranged the dàdiǎn extracts (which had been redistributed by rhyme and so lost their mén affiliation) provisionally by historical period. The result is the 8-juàn WYG text, which is the basis of all modern editions.

The standard modern edition is Zhōu Xūnchū’s 周勛初 Táng yǔlín jiàozhèng 唐語林校證 (Zhōnghuá 1987, TángSòng shǐliào bǐjì cóngkān), which reconstructs the work against the Xīn / Jiù Tángshū and the parallel Táng anecdote literature; this is the edition Wilkinson and A Sung Bibliography recommend.

Translations and research

  • Qin, Amelia Ying. 2018. “An Introductory Study of the Tang yulin.” Tsing Hua Journal of Chinese Studies 新亞學報 48.2: 419–56. The principal English-language scholarly treatment.
  • Zhōu Xūn-chū 周勛初, ed. 1987. Táng yǔlín jiào-zhèng 唐語林校證. 2 vols. Běi-jīng: Zhōng-huá shū-jú. Standard critical edition; reconstructs sources, corrects the Sìkù re-ordering, identifies parallel passages in Xīn / Jiù Táng shū and the surviving Táng xiǎo-shuō.
  • Hartman, Charles, and Anna M. Shields, eds. The Many Faces of Lord Niu: Niu Sengru (780–848) and the Politics of Memory in Late Medieval China. Mines Táng yǔlín extensively for late-Táng faction-politics anecdotes.
  • A Sung Bibliography, ed. Yves Hervouet (Hong Kong: Chinese University Press, 1978), 97–98, entry by Hervouet himself, is the standard Western-language descriptive notice.
  • No complete European-language translation has been located.

Other points of interest

The Táng yǔlín is the locus classicus of many famous Táng anecdotes about Lǐ Bái 李白, Dù Fǔ 杜甫, Bái Jūyì 白居易, Hán Yù 韓愈, Liǔ Zōngyuán 柳宗元, Niú Sēngrú 牛僧孺, and Lǐ Déyù 李德裕 — including the well-known account of Bái Jūyì reading his poems to an old peasant woman to test their accessibility. Because Wáng’s excerpting preserves Táng material from books since lost (notable losses include Lǐ Zhào’s 李肇 Hànlín zhì 翰林志 in fuller form, and large parts of the Yǔlín 語林 of Péi Qǐ 裴啟 lineage), the work functions in practice as a Táng-source repository, and modern Táng-studies scholarship treats it as a primary rather than secondary source where the original xiǎoshuō is lost.

The fifty-two-mén expansion of the Shìshuō’s thirty-five headings is itself an important moment in the history of the Shìshuō genre, demonstrating how the Shìshuō model could be adapted to a different dynasty’s social and intellectual world; subsequent imitations (e.g., the Míng Héshì yǔlín 何氏語林, the Qīng Sòng yàn 宋豔) follow this expanded heading-system more often than the original 35.

Graph View

Backlinks