Shīlín guǎng jì 詩林廣記

The Broad Records of the Forest of Poetry by 蔡正孫 (撰)

About the work

The Shīlín guǎng jì 詩林廣記, in two parts of ten juǎn each — Qián jí 前集 and Hòu jí 後集, twenty juǎn total — is a major Yuán-period anthology-cum-criticism of Chinese poetry compiled by Cài Zhèngsūn 蔡正孫 ( Cuìrán 粹然, hào Méngzhāi yěyì 蒙齋野逸), a Southern Sòng loyalist and follower of Xiè Fǎngdé 謝枋得. The author’s preface is dated to the cyclical yǐchǒu year — pointedly without any reign-title — which the Sìkù editors identify as Yuán Zhìyuán 至元 26 (1289), thirteen years after the fall of the Southern Sòng. The format is now famous: each entry takes a poet, gives the poem in full, and then drops two grid-spaces and appends the assembled critical comments from earlier shīhuà. The Qián jí runs from Táo Qián 陶潛 to Yuán Zhěn 元稹 (Yuán Wēizhī 微之), twenty-four poets, with a fù lù 附錄 in juǎn 9 (Xuē Néng 薛能 and two others) and a further fù lù in juǎn 10 (Xuē Dàohéng 薛道衡 and four others). The Hòu jí runs from Ōuyáng Xiū 歐陽修 to Liú Bān 劉攽, twenty-eight poets, and stops short of the Southern Sòng. Cài’s preface promised a Xù jí 續集 with the remaining poets; no Xù jí survives, and probably none was ever printed. The two extant parts together amount to one of the most influential pre-Míng compilations of pre-Sòng / Northern Sòng verse criticism. The Qīng Sòngshī jì shì 宋詩紀事 of Lì È 厲鶚 (1746) consciously adopted Cài’s procedure — and through that route Cài’s format entered modern scholarship.

Tiyao

Shīlín guǎng jì, Qián jí 10 juǎn, Hòu jí 10 juǎn. By Cài Zhèngsūn of the Sòng. Zhèngsūn’s was Cuìrán; his self-style was Méngzhāi yěyì. The book is prefaced by Cài’s own preface, dated “the year-star at túwéi chìfènruò” (a poetic gānzhī expression = jǐchǒu 己丑 / 已丑); a glance through the entry on Huáng Tíngjiān’s 黃庭堅 poem to Sū Zhé 蘇轍 (which quotes Xióng Hé 熊禾) fixes the year as Yuán Shìzǔ Zhìyuán 26 — the Sòng had fallen ten years before, hence Cài’s omission of the reign-title (his “writing only the cyclical date”, in the manner of Táo Qián). In the appendices to Xiè Fǎngdé’s collection, in the zèngxíng (parting-poem) section, there is a poem by Zhèngsūn — evidently the same man.

The Qián jí covers twenty-four poets from Táo Qián to Yuán Wēizhī. Juǎn 9 carries an appendix on Xuē Néng and two others; juǎn 10 carries an appendix on Xuē Dàohéng and four others. The Hòu jí covers twenty-eight poets from Ōuyáng Xiū to Liú Bān, ending with the Northern Sòng. At the end of the table of contents he notes that “those not yet included in the selection will appear in a forthcoming Xù jí” — but no Xù jí is now to be seen.

Both arrange the material by poet, with critical comments arranged under each poem. Each poem is given in full first; the quoted critical observations are then placed two spaces below the poem, set out one after another. Its format sits between a zǒngjí and a shīhuà. The present dynasty’s Lì È, in compiling the Sòngshī jì shì, in fact adopts this procedure. The difference is only that for any poem where the present book has no critical comment or kǎozhèng to attach, that poem is not included — so the book does not, like Lì È’s, simply transcribe poems indiscriminately, on the model of the Tángshī jì shì. That is a small but real distinction.

Abstract

The Shīlín guǎng jì is the most influential single contribution of the SòngYuán transitional generation to the shīhuà tradition. As an anthology — it preserves the full text of every poem it discusses — and as a critical archive — it gathers, for each poem, the assembled comments of earlier shīhuà, often with attribution — it created a hybrid format that decisively shaped subsequent YuánMíngQīng poetic anthologization. Lì È’s Sòngshī jì shì (1746), the standard Qīng anthology of Sòng poetry, is its direct descendant.

Cài’s biographical situation is essential to the book. He was a yímín of the Sòng — and not just any yímín but a personal disciple of Xiè Fǎngdé 謝枋得, the most prominent Southern Sòng loyalist martyr of the post-1276 generation. The connection is fixed by the Diéshān jí (the Xiè Fǎngdé collection), in whose zèngxíng section appears a parting poem from Cài Zhèngsūn to his teacher. Cài’s pointed decision to date the preface in cyclical gānzhī alone — yǐchǒu 已丑 / 己丑 — and to refer to himself as “Méngzhāi yěyì” (the Hidden Idler of Méng Studio) places him in the Táo Qián / Tóngyīn shī huà lineage of loyalist refusal: cyclical dating after the fall of the dynasty, refusal of office, withdrawal to compilation. The catalog’s assignment to “宋” reflects not the calendar but Cài’s self-conception. The work itself was completed under Yuán Zhìyuán 26 (1289), so a single composition date is appropriate; this is followed here.

The textual transmission begins with a Sòng manuscript that Cài himself describes in the preface as compiled with “no library to consult” — a self-deprecation given the book’s range (more than fifty sources are quoted). The Yuán printing tradition survives in fragments; the Míng Jiājìng and Wànlì reprintings (notably the Yáng Lián 楊鎌 edition) became the standard. The Sìkù edition derives from Jì Yún 紀昀’s family copy and is the basis of all subsequent re-editions; the modern punctuated Lìdài shīhuà recension (Hé Wénhuàn 何文煥) and the Sòng shīhuà quán biān (Wú Wénzhì 吳文治, 1998) both rest on it. As a quotational archive the Shīlín guǎng jì preserves substantial fragments of Sòng shīhuà that survive in no other form, most notably the Wáng Zhífāng shīhuà 王直方詩話, the Lěngzhāi yèhuà 冷齋夜話 (Huì Hóng 惠洪), and several Northern Sòng critical works lost in the Mongol conquest.

The work belongs to the same critical-anthological wave as Hé Xīwèn’s 何谿汶 Zhúzhuāng shīhuà (KR4i0041); the Sìkù editors compare the two explicitly as the leading examples of the hybrid form, with the difference that Hé places critical comments before the poem, Cài after.

Translations and research

  • Zhāng Bóběi 張伯偉, Zhōng-guó gǔ-dài wén-xué pī-píng fāng-fǎ yán-jiū 中國古代文學批評方法研究 (Zhōnghuá, 2002) — extended discussion of the Shī-lín guǎng jì and the late-Sòng/early-Yuán anthology-criticism form.
  • Cài Zhèn-chǔ 蔡鎮楚, Shī huà xué 詩話學 (Húnán jiào-yù, 1990).
  • Guō Shào-yú 郭紹虞, Sòng shī-huà jí yì 宋詩話輯佚 (Zhōnghuá, 1980 reprint) — uses Shī-lín guǎng jì as a principal recovery source.
  • Modern punctuated edition: Shī-lín guǎng jì, ed. Cháng Zhèn-guó 常振國 and Jiǎng Yùn 絳雲 (Rénmín wén-xué, 1982; rev. 1987).
  • Zhāng Hóng-shēng 張宏生, Sòng shī: Róng-huì yǔ kāi-tuò 宋詩:融匯與開拓 (Shàng-hǎi gǔ-jí, 2001) — discusses Cài’s selection criteria as evidence for late-Sòng / early-Yuán reading tastes.

Other points of interest

The work is one of the most celebrated literary-historical instances of yímín refusal. Cài’s adherence to Xiè Fǎngdé — Xiè starved himself rather than serve the Yuán and is the model of Sòng-loyalist martyrdom — places the Shīlín guǎng jì in a direct ideological line: Cài’s “anthology of the irrecoverable past” can be read as a literary-critical analogue of his teacher’s hunger strike. The pointed omission of the Yuán reign-title from the preface, on the Táo Qián model (Táo’s own post-Jìn dating in cyclical years alone is the precedent Cài invokes by including Táo as the Qián jí’s opening poet), is the principal textual signal of this loyalist stance. The work has thus been read both as a literary-critical innovation and as a yímín document.