唐百家詩選

Táng bǎijiā shī xuǎn 唐百家詩選

Selected Poems of One Hundred Táng Poets by 王安石

About the work

A twenty-juǎn anthology of approximately 1,250 poems by 108 Táng poets (the WYG count is 1,262; Cháo Gōngwǔ’s Dúshū zhì reports 1,246), traditionally attributed to Wáng Ānshí 王安石 (1021–1086) and produced ca. 1058–1064 when Wáng and Sòng Mǐnqiú 宋敏求 (1019–1079) were colleagues at the Sānsī pànguān office. The compilation is generally agreed to have been a collaborative project: Sòng Mǐnqiú made available the 108 Táng-poet collections in his private library; Wáng Ānshí made the critical selection; the staff copyists assembled the fair copy. The book is notable, and persistently controversial, for excluding the very greatest Táng poets — there are no poems by Lǐ Bái, Dù Fǔ, or Hán Yù, and minimal representation of Wáng Wéi, Wéi Yìngwù, Yuán Zhěn, Bái Jūyì, Liú Yǔxī, Liǔ Zōngyuán, Mèng Jiāo, and Zhāng Jí. The principle of selection, stated in Wáng Ānshí’s own preface, was to anthologise the secondary names (“those whose collections are little circulating”) — an editorial decision that the SKQS editors and others have variously praised and disparaged.

Tiyao

Your servants respectfully submit: the Táng bǎijiā shī xuǎn in twenty juǎn. The old form bears the title of “compiled by Wáng Ānshí of the Sòng.” Its inclusion-and-exclusion are entirely beyond reason. Since the Sòng, those who doubt the attribution are not few, and the various explanations of how this could be by Wáng are not few either. Most still hold it to be by him. Cháo Gōngwǔ’s Dúshū zhì alone says: Táng bǎijiā shī xuǎn in twenty juǎn was edited by Sòng Mǐnqiú (zì Cìdào). Cìdào was Sānsī pànguān and once took the family-stored 108 Táng-poet collections and chose the choicest, 1,246 pieces in all, compiling them; Wáng Jièfǔ saw it and made his own further selection, writing on it: “He who would observe Táng poetry may stop here.” The world then took it as Jièfǔ’s. His view is distinctively at variance with all others.

The Dúshū zhì was written in the early Southern Sòng, not long after Ānshí’s time; further, the Cháo family was a learned house from Yuányòu on, transmitting traditions, so his testimony must have a foundation. Shào Bó’s Wénjiàn hòulù cites Cháo Yuēzhī’s words: Wáng Jīnggōng and Sòng Cìdào were colleagues at the Qúnmùsī pànguān office. Cìdào’s house held many Táng-poet collections. Jīnggōng took them and marked the choicest with paper-tags, instructing the clerks to copy; the clerks, weary of the writing, would move the tags from long poems they had marked to short poems they had not, and Jīnggōng, careless by nature, did not check again. So what the world now calls the “Jīnggōng selection” of the Táng bǎijiā shī xuǎn is in fact the clerks’ selection. This account differs again from Cháo Gōngwǔ’s. If Cháo Yuēzhī had made the clerk-selection claim, surely Gōngwǔ would know of it. Yet Zhōu Huī’s Qīngbō zázhì has the same story; Zhōu’s great-grandfather was Wáng Ānshí’s cousin, so Zhōu’s stance is generally pro-Wáng — perhaps the story was put about by Ānshí’s faction to defend him from public displeasure at the anthology, and Cháo Yuēzhī and Shào Bó simply repeated it without checking.

This text is the Qiándào (Sòng) printing of Ní Zhòngchuán 倪仲傳; before it is Ní’s preface; the book had been out of circulation for a long time. In the Kāngxī of our dynasty, Sòng Luò of Shāngqiū bought a fragmentary copy of 8 juǎn and printed it; later got the complete book and reprinted, so the 20 juǎn are again whole. At the time some doubted its authenticity. Yán Ruòjǔ 閻若璩 cited Gāo Bǐng’s Tángshī pǐnhuì KR4h0095 (which gives Mínghuáng’s Zǎo dù Púguān as the opening piece) and Chén Zhènsūn’s Shūlù jiětí (which notes the absence of LǐDùHán and even Wáng Wéi, Wéi Yìngwù, YuánBái, LiúLiǔ, Mèng Jiāo, Zhāng Jí) — to prove it is genuine. The fragmentary version had lost the Wáng Ānshí preface; Yán Ruòjǔ supplied it from the Línchuān jí — printed in full in his Qiánqiū zhájì.

Only — the present book records 1,262 pieces, 16 more than Cháo’s number. Yán did not comment; or perhaps a copyist of the Dúshū zhì wrote 6/2 for 4/6. Again, in Wáng Chānglíng’s “Chū sài,” all texts read ruò shǐ Lóngchéng fēijiàng zài 若使龍城飛將在, but this text reads Lúchéng fēijiàng 盧城飛將. Yán cites the Táng Píngzhōu governance of Lúlóng xiàn as evidence — but poets often misuse names, and one need not defend it. The whole three hundred years of Táng have not one other man who calls Lúlóng Lúchéng: why should Chānglíng alone fabricate a place name? Yán’s excessive reverence for the Sòng-block text is a flaw that should not be made standard. Reverently submitted, third month of Qiánlóng 43 (1778). Editor-in-Chief Jǐ Yún, Lù Xīxióng, Sūn Shìyì. General Collator Lù Fèichí.

The Ní Zhòngchuán preface of Qiándào jǐchǒu (1169): “Sounds that are subtle are hard to appreciate; tunes that are high have few who can match — naturally the Táng bǎijiā shī xuǎn has fallen out of the world. I from my ruòguàn studied under Xiāngxī xiānsheng [Tàidù] and found this collection in his hidden library, loved how it draws out the clearer-and-more-ancient, the diǎnlì (canonical-and-beautiful) Táng poetry — pure but not lewd; whatever Táng poet had a striking line is gathered here…” And Wáng Ānshí’s own preface: “I served as Sānsī pànguān with Sòng Cìdào; Cìdào produced his family-stored one hundred and odd Táng-poet collections and asked me to choose the finest. Cìdào named the result Bǎijiā shī xuǎn. We spent days at it…”

Abstract

Date: by the prefaces — Wáng Ānshí and Sòng Mǐnqiú were both at the Sānsī pànguān office ca. 1058–1064; the compilation belongs to that period. (Cháo Yuēzhī’s “Qúnmùsī pànguān” attribution may be confusion or a different office.) The book was printed during the Qiándào era (1165–1173) by Ní Zhòngchuán and substantially circulated thereafter, but was driven out of circulation by the more comprehensive YuánMíng anthologies; rediscovered and reprinted under Kāngxī by Sòng Luò.

The historiographical significance is twofold: (1) the selection criterion — anthologise the secondary canon — is unique among major Táng anthologies and gives the Bǎijiā shī xuǎn its principal documentary value: it is the single richest Sòng-period source for the second-tier Táng poets whose individual collections are now lost or fragmentary. Approximately twenty of the 108 poets in the anthology have no surviving individual collection apart from this anthology. (2) The persistent ambiguity of the editorship — Wáng vs. Sòng Mǐnqiú vs. the clerks — is a classic problem in the Northern-Sòng critical scholarship, and the SKQS editors’ careful re-examination is exemplary.

In modern scholarship the anthology is read primarily as Sòng critical practice: Wáng Ānshí’s exclusion of LǐDùHán is taken as a programmatic statement about which Táng tradition Sòng poetics should inherit — namely, the Wáng Wéi / Mèng Hàorán “natural” lineage rather than the LǐDù grand canon. This reading is supported by Wáng Ānshí’s own biéjí poetics and was extended by his Northern-Sòng followers (Cháo Bǔzhī, Huáng Tíngjiān).

Translations and research

• Charles Hartman, “Wang Anshi’s Tang Anthology,” Bulletin of Sung-Yuan Studies 19 (1987): 1–20.
• Wáng Shuǐ-zhào 王水照, Wáng Ān-shí xuéshù sī-xiǎng yánjiū 王安石學術思想研究 — section on the Táng bǎi-jiā shī xuǎn.
• Fù Xuán-cóng 傅璇琮, Táng-rén xuǎn Táng-shī xīn biān (Xī’ān: Shǎnxī rénmín, 1996).
• Peter Bol, “This Culture of Ours” (Stanford, 1992), brief discussion of Wáng’s anthology in his account of Northern-Sòng intellectual reform.

Other points of interest

The “Lúchéng / Lóngchéng” variant in Wáng Chānglíng’s Chūsài, debated by Yán Ruòjǔ and the SKQS editors, is one of the most-cited textual cruces in Táng anthology criticism. The SKQS editors’ arguing against Yán Ruòjǔ — that Lóngchéng must be original and Lúchéng a Sòng-printed scribal error — is now the orthodox modern view.

Links

• Wilkinson, Chinese History: A New Manual §30.3.1, §31.4.
• ctext
• Wikipedia, “Wang Anshi”