藏海詩話

Cánghǎi shīhuà 藏海詩話

Cáng-hǎi’s Remarks on Poetry by 吳可 (撰)

About the work

The Cánghǎi shīhuà 藏海詩話, in one juǎn, survives only through extraction from the Yǒnglè dàdiǎn 永樂大典 — the Míng compendium of 1408 — into the Qīng Sìkù edition; the work was anonymous in the Yǒnglè dàdiǎn recension and the Sìkù editors reconstructed the authorship by internal evidence. They argue with cogency that it is by Wú Kě 吳可 (hào Cánghǎi 藏海, fl. ca. 1119–1170s), the author of the Cánghǎi jūshì jí 藏海居士集 KR4d0183: the jí (collected works) and the shīhuà share the studio-name, the same Línchuān Tóng Démǐn 童德敏 acquaintance, the same Yán Lǔgōng shrine reference, the same Wáng Shēn 王詵 Chūnjiāng tú dedication-poem, and the same chronological window down to the 1170s. The Sìkù editors’ attribution is universally accepted in modern scholarship.

The book opens a Sòng-poetic outlook between Jiāngxī 江西 and Chán (Zen) meditation: Wú’s central theoretical formulations — shī yǐ yòngyì wéi zhǔ ér fù zhī yǐ huálì 詩以用意爲主而附之以華麗 (“poetry takes intent-of-mind as its master and adorns it with brilliance”), xué shī dāng yǐ Dù wéi tǐ, yǐ Sū Huáng wéi yòng 學詩當以杜爲體以蘇黃爲用 (“to learn poetry, take Dù Fǔ as substance, Sū Shì and Huáng Tíngjiān as function”), and the famous distinction between yǒuxíng zhī bìng 有形之病 and wúxíng zhī bìng 無形之病 (“the visible fault and the invisible fault”) — are among the most precisely-stated of Sòng poetic doctrines and were a direct influence on Yán Yǔ’s 嚴羽 Cānglàng shīhuà KR4i0035 a century later.

Tiyao

Case: the Cánghǎi shīhuà is found in the Yǒnglè dàdiǎn; it does not give the compiler’s name. From the Míng dynasty on, no catalog has registered it. Examining the Yǒnglè dàdiǎn: a Cánghǎi jūshì jí by Wú Kě of the Sòng was already gathered and edited and separately catalogued. In that jí there is a Tí Wáng Shēn Chūnjiāng tú (Wáng Shēn = Wáng Jìnqīng); and there is much of poetry-discussion exchanges with Hán Jū 韓駒. The Xuānhé and Zhènghé (1111–1125, 1119–1125) years cited, and the Jiànyán (1127ff.) initial flight south through Húnán and Lǐngnán — match the date-coordinates of this present book, in which the closing entry refers to “from Yuányòu until now, sixty-some years”. The match is exact: this is presumably Wú’s work.

Wú’s discussion of poetry often runs into deliberate bùliǎoliǎo yǔ 不了了語 (“indeterminate sayings”) — somewhat Chán-style sword-point retort, not quite free of bad habit. Other passages — citing Xú Fǔ 徐俯’s reading of Dù Fǔ’s Tiānjí màn qīng sī 天棘蔓靑絲 (“the tiānjí trails green silk”) as “seeing the willow and remembering the horse” — are too tortured. He twists yú 渝 (transgress) into lún 淪 (sunk), and corrupts Guǎng yǎ 廣雅 into Ěr yǎ 爾雅 — slight slips.

But he himself saw the Yuányòu old generation; his learning has somewhere to anchor. What he says — shī yǐ yòngyì wéi zhǔ ér fù zhī yǐ huálì, nìng duì bù gōng, bù kě shǐ qì ruò (“poetry takes intent-of-mind as master and is adorned with brilliance; better to break the parallelism than to let the qì slacken”) — corrects the nóngyàn 穠豔 (heavy adornment) error of the Xīkūn 西崑 school. And — fán kàn shī, xū shì yī piān lì yì, nǎi yǒu guīsù chù (“in reading poetry, the whole piece must establish its intent, only then is there somewhere to come back to”) — and — xué shī dāng yǐ Dù wéi tǐ, yǐ Sū Huáng wéi yòng; Dù zhī miàochù cáng yú nèi, Sū Huáng zhī miàochù fā yú wài (“Dù’s wonder hides inside; SūHuáng’s wonder issues outside”) — and — juéjù rú xiǎojiā shì, jùzhōng zháo dàjiā shì bùdé (“the quatrain is a small-house affair; one cannot install great-house affairs in it”); citing Shāngǔ (= Huáng Tíngjiān)‘s Crab poem using hǔzhēng 虎爭 and zhījiě 支解 — the affair is too big and does not belong in the verse) — and — qīyán lǜshī jí nán zuò, gài yìdé sú, suǒyǐ Shāngǔ bié wéi yī tǐ (“the seven-character regulated lǜshī is exceedingly hard to make; it easily becomes vulgar; this is why Shāngǔ made it into a separate style of its own”): all profoundly true. His remarks on the yǒuxíng zhī bìng (the visible fault) and the wúxíng zhī bìng (the invisible fault) are penetrating to a fine grain.

The rest of his critical readings and verifications are also worth keeping. Yet Hú Zǎi 胡仔’s Tiáoxī yúyǐn cónghuà KR4i0029 and Wèi Qìngzhī 魏慶之’s Shīrén yùxiè KR4i0036 — copious as they are — both fail to extract from it. So Wú was already obscure in the Sòng. The book is therefore properly brought forward and exhibited, for the use of those who discuss this art.

Abstract

The Cánghǎi shīhuà is the personal shīhuà of an obscure but theoretically sophisticated early-Southern-Sòng poet — Wú Kě 吳可, who had been an official in Biànjīng at the close of the Northern Sòng, then a refugee through Húnán and Lǐngnán during the Jiànyán flight, and a quiet literary correspondent of Hán Jū 韓駒, Wáng Ānzhōng 王安中, Mǐ Yǒurén 米友仁, and Wáng Shēn 王詵. The Sìkù editors reconstructed his authorship from the Yǒnglè dàdiǎn recension and the parallel transmission of his Cánghǎi jūshì jí KR4d0183; their attribution is decisive. Internal evidence — the closing reference to “from Yuányòu until now, sixty-some years” — places the book’s composition or final assembly in the 1130s–1140s window, though some entries are clearly earlier; the work was demonstrably still circulating in the Qiándào / Chúnxī 1170s, since it cites a Hóng Mài 洪邁 Róngzhāi sānbǐ 容齋三筆 figure (Tóng Démǐn 童德敏).

The book is doctrinally distinctive for three formulations that became standard reference points in subsequent Sòng poetics:

1. The intent-and-adornment doctrine. “Poetry takes intent-of-mind (yòngyì) as master, and adorns it with brilliance (huálì); better to break the parallelism than to let the qì go weak.” This is a direct counter-formulation to the Xīkūn 西崑 school’s elevation of ornament over substance.

2. The Dù-as-substance, Sū-Huáng-as-function doctrine. “In learning poetry, take Dù as tǐ (substance), Sū and Huáng as yòng (function). Dù’s wonder hides inside; SūHuáng’s wonder issues outside.” This sets up the canonical Sòng triad (DùSūHuáng) on a Lǐxué substance/function template — a remarkable application of Lǐxué ontology to poetic theory.

3. The visible-vs-invisible-fault doctrine (yǒuxíng zhī bìng 有形之病 / wúxíng zhī bìng 無形之病). The visible fault — broken parallelism, wrong allusion, weak word — can be diagnosed and corrected; the invisible fault — slackness of qì, lack of intent, lack of a “place to come back to” — eludes mechanical fix and is the true measure of poetic mastery. This formulation anticipates Yán Yǔ’s 嚴羽 miàowù 妙悟 (subtle awakening) doctrine in the Cānglàng shīhuà KR4i0035 of the early thirteenth century, and Yán’s debt to Wú is direct (despite the latter’s neglect by Hú Zǎi and Wèi Qìngzhī).

The work’s Chán (Zen) flavor — recognized but criticized by the Sìkù editors as bùliǎoliǎo yǔ — is one of the earliest sustained applications of Chán discourse to shīhuà writing, and is again a direct predecessor of Yán Yǔ’s yǐ Chán yù shī 以禪喻詩 (“explicating poetry by Chán”) doctrine.

Transmission: the work survives only through the Yǒnglè dàdiǎn recension extracted into the Sìkù. It was lost to all Sòng shīhuà compendia; its rediscovery by the Sìkù editors restored an important but missing link in the genealogy between Sòng Jiāngxī doctrine and Yán Yǔ’s late-Sòng synthesis.

Translations and research

• Guō Shào-yú 郭紹虞, Sòng shī-huà jí-yì 宋詩話輯佚 (Zhōnghuá, 1980) — includes the Cáng-hǎi shī-huà; principal modern reference.
• Cài Zhèn-chǔ 蔡鎮楚, Zhōng-guó shī-huà shǐ 中國詩話史 (Hú-nán Wén-yì, 1988) — discusses Wú Kě’s anticipation of Yán Yǔ.
• Mò Lì-fēng 莫礪鋒, Jiāng-xī shī-pài yán-jiū 江西詩派研究 (Qí-Lǔ shū-shè, 1986).
• Zhāng Bǎo-quán 張伯偉, Chán yǔ shī-xué 禪與詩學 (Hángzhōu: Zhè-jiāng Rén-mín, 1992) — places Wú in the Chán-poetry crossover.
• Stephen Owen, Readings in Chinese Literary Thought (Harvard, 1992).
• Hé Wén-huàn 何文煥, ed., Lì-dài shī-huà 歷代詩話 (rpt. Zhōnghuá, 1981).

Other points of interest

The Cánghǎi shīhuà’s anticipation of Yán Yǔ’s Cānglàng shīhuà KR4i0035 — particularly the yǐ Chán yù shī 以禪喻詩 approach and the yǒuxíng / wúxíng zhī bìng doctrine — is one of the clearer cases of a missing-link text recovered by the Sìkù editors from the Yǒnglè dàdiǎn. Without the Cánghǎi shīhuà, Yán Yǔ’s synthesis would look more discontinuous than it actually was.

— The fact that the two great Sòng shīhuà compendia (Hú Zǎi KR4i0029 and Wèi Qìngzhī KR4i0036) failed to cite Wú Kě — qí chénhuì yǐ jiǔ “his obscurity was already long” — is a useful caution about the limits of even the most comprehensive contemporary anthologies as evidence for the shīhuà corpus of their period.

Links

• Wilkinson, Chinese History: A New Manual §30.5.
• Kyoto Zinbun Sìkù tíyào
• Guō Shàoyú, Sòng shīhuà jíyì (Zhōnghuá, 1980).